paper_id string | title string | paper_url string | authors list | type string | primary_area string | abstract large_string | keywords list | TL;DR large_string | submission_number int64 | arxiv_id string | arxiv_id_source string |
|---|---|---|---|---|---|---|---|---|---|---|---|
VD0nXUG5Qk | RoboTube: Learning Household Manipulation from Human Videos with Simulated Twin Environments | https://openreview.net/forum?id=VD0nXUG5Qk | [
"Haoyu Xiong",
"Haoyuan Fu",
"Jieyi Zhang",
"Chen Bao",
"Qiang Zhang",
"Yongxi Huang",
"Wenqiang Xu",
"Animesh Garg",
"Cewu Lu"
] | Oral | null | We aim to build a useful, reproducible, democratized benchmark for learning household robotic manipulation from human videos. To realize this goal, a diverse, high-quality human video dataset curated specifically for robots is desired. To evaluate the learning progress, a simulated twin environment that resembles the a... | [
"Learning from Videos",
"Video Demonstration Dataset",
"Real2Sim",
"Self-supervised Reward Learning",
"Robotic Simulation Benchmark"
] | null | 1 | null | null |
SM70KHTBG-0 | Vision-based Uneven BEV Representation Learning with Polar Rasterization and Surface Estimation | https://openreview.net/forum?id=SM70KHTBG-0 | [
"Zhi Liu",
"Shaoyu Chen",
"Xiaojie Guo",
"Xinggang Wang",
"Tianheng Cheng",
"Hongmei Zhu",
"Qian Zhang",
"Wenyu Liu",
"Yi Zhang"
] | Poster | null | In this work, we propose PolarBEV for vision-based uneven BEV representation learning. To adapt to the foreshortening effect of camera imaging, we rasterize the BEV space both angularly and radially, and introduce polar embedding decomposition to model the associations among polar grids. Polar grids are rearranged to... | [
"Polar Rasterization",
"Surface Estimation",
"BEV Segmentation"
] | We propose PolarBEV for vision-based uneven BEV representation learning. | 3 | 2207.01878 | title_snapshot |
VHia4Cint7 | HERD: Continuous Human-to-Robot Evolution for Learning from Human Demonstration | https://openreview.net/forum?id=VHia4Cint7 | [
"Xingyu Liu",
"Deepak Pathak",
"Kris M. Kitani"
] | Poster | null | The ability to learn from human demonstration endows robots with the ability to automate various tasks. However, directly learning from human demonstration is challenging since the structure of the human hand can be very different from the desired robot gripper. In this work, we show that manipulation skills can be tra... | [
"learning from human demonstration",
"imitation learning",
"curriculum learning",
"transfer learning",
"robotic manipulation"
] | A method for learning from human demonstration using continuous robot evolution. | 5 | 2212.04359 | title_snapshot |
80vpxjt3vq | PlanT: Explainable Planning Transformers via Object-Level Representations | https://openreview.net/forum?id=80vpxjt3vq | [
"Katrin Renz",
"Kashyap Chitta",
"Otniel-Bogdan Mercea",
"A. Sophia Koepke",
"Zeynep Akata",
"Andreas Geiger"
] | Poster | null | Planning an optimal route in a complex environment requires efficient reasoning about the surrounding scene. While human drivers prioritize important objects and ignore details not relevant to the decision, learning-based planners typically extract features from dense, high-dimensional grid representations containing a... | [
"Autonomous Driving",
"Transformers",
"Explainability"
] | We propose PlanT, a state-of-the-art planner for self-driving based on object-level representations and a transformer architecture which can explain its decisions by identifying the most relevant object. | 16 | 2210.14222 | title_snapshot |
WkbvHMQL6uJ | Where To Start? Transferring Simple Skills to Complex Environments | https://openreview.net/forum?id=WkbvHMQL6uJ | [
"Vitalis Vosylius",
"Edward Johns"
] | Poster | null | Robot learning provides a number of ways to teach robots simple skills, such as grasping. However, these skills are usually trained in open, clutter-free environments, and therefore would likely cause undesirable collisions in more complex, cluttered environments. In this work, we introduce an affordance model based on... | [
"Robot manipulation",
"planning",
"obstacle avoidance"
] | null | 17 | 2212.06111 | title_snapshot |
52uzsIGV32_ | Efficient and Stable Off-policy Training via Behavior-aware Evolutionary Learning | https://openreview.net/forum?id=52uzsIGV32_ | [
"Maiyue Chen",
"Guangyi He"
] | Poster | null | Applying reinforcement learning (RL) algorithms to real-world continuos control problems faces many challenges in terms of sample efficiency, stability and exploration. Off-policy RL algorithms show great sample efficiency but can be unstable to train and require effective exploration techniques for sparse reward envir... | [
"Continuos control",
"Reinforcement learning",
"Evolution strategies"
] | An evolutionary training framework for off-policy reinforcement learning inspired by evolution strategies (ES) called Behavior-aware Evolutionary Learning (BEL). | 19 | null | null |
UW5A3SweAH | LM-Nav: Robotic Navigation with Large Pre-Trained Models of Language, Vision, and Action | https://openreview.net/forum?id=UW5A3SweAH | [
"Dhruv Shah",
"Błażej Osiński",
"brian ichter",
"Sergey Levine"
] | Poster | null | Goal-conditioned policies for robotic navigation can be trained on large, unannotated datasets, providing for good generalization to real-world settings. However, particularly in vision-based settings where specifying goals requires an image, this makes for an unnatural interface. Language provides a more convenient mo... | [
"instruction following",
"language models",
"vision-based navigation"
] | We can utilize pre-trained models of images and language to provide a textual interface to visual navigation models, enabling zero-shot instruction following robots in the real world! | 21 | 2207.04429 | title_snapshot |
A5l7wE2uqtM | BusyBot: Learning to Interact, Reason, and Plan in a BusyBoard Environment | https://openreview.net/forum?id=A5l7wE2uqtM | [
"Zeyi Liu",
"Zhenjia Xu",
"Shuran Song"
] | Poster | null | We introduce BusyBoard, a toy-inspired robot learning environment that leverages a diverse set of articulated objects and inter-object functional relations to provide rich visual feedback for robot interactions. Based on this environment, we introduce a learning framework, BusyBot, which allows an agent to jointly acqu... | [
"Manipulation",
"Scene Understanding",
"Learning Environment"
] | We propose a toy-inspired relational environment, BusyBoard, and a learning framework, BusyBot, for embodied AI agents to acquire interaction, reasoning, and planning skills. | 25 | 2207.08192 | title_snapshot |
sK2aWU7X9b8 | Training Robots to Evaluate Robots: Example-Based Interactive Reward Functions for Policy Learning | https://openreview.net/forum?id=sK2aWU7X9b8 | [
"Kun Huang",
"Edward S. Hu",
"Dinesh Jayaraman"
] | Oral | null | Physical interactions can often help reveal information that is not readily apparent. For example, we may tug at a table leg to evaluate whether it is built well, or turn a water bottle upside down to check that it is watertight. We propose to train robots to acquire such interactive behaviors automatically, for the pu... | [
"Reinforcement Learning",
"Interactive Perception",
"Task Specification"
] | We present LIRF, a framework for conveniently training example-based interactive robot policies to evaluate robot task policies, in order both to provide rewards to train them, and to verify their execution in partially observed tasks. | 26 | 2212.08961 | title_snapshot |
52c5e73SlS2 | Walk These Ways: Tuning Robot Control for Generalization with Multiplicity of Behavior | https://openreview.net/forum?id=52c5e73SlS2 | [
"Gabriel B. Margolis",
"Pulkit Agrawal"
] | Oral | null | Learned locomotion policies can rapidly adapt to diverse environments similar to those experienced during training but lack a mechanism for fast tuning when they fail in an out-of-distribution test environment. This necessitates a slow and iterative cycle of reward and environment redesign to achieve good performance o... | [
"Locomotion",
"Reinforcement Learning",
"Task Specification"
] | Learning quadrupedal locomotion with an expanded task specification enables online tuning of a small quadruped to out-of-distribution environments and tasks. | 27 | 2212.03238 | title_snapshot |
OzP68WT3UA | NeuralGrasps: Learning Implicit Representations for Grasps of Multiple Robotic Hands | https://openreview.net/forum?id=OzP68WT3UA | [
"Ninad Khargonkar",
"Neil Song",
"Zesheng Xu",
"B Prabhakaran",
"Yu Xiang"
] | Poster | null | We introduce a neural implicit representation for grasps of objects from multiple robotic hands. Different grasps across multiple robotic hands are encoded into a shared latent space. Each latent vector is learned to decode to the 3D shape of an object and the 3D shape of a robotic hand in a grasping pose in terms of t... | [
"Robot Grasping",
"Neural Implicit Representations",
"Grasp Transfer",
"Grasping Contact Modeling",
"6D Object Pose Estimation"
] | Learning implicit representations for multiple robot hands in a common encoding space for flexible grasp transfer. | 29 | 2207.02959 | title_snapshot |
d-JYso87y6s | Frame Mining: a Free Lunch for Learning Robotic Manipulation from 3D Point Clouds | https://openreview.net/forum?id=d-JYso87y6s | [
"Minghua Liu",
"Xuanlin Li",
"Zhan Ling",
"Yangyan Li",
"Hao Su"
] | Poster | null | We study how choices of input point cloud coordinate frames impact learning of manipulation skills from 3D point clouds. There exist a variety of coordinate frame choices to normalize captured robot-object-interaction point clouds. We find that different frames have a profound effect on agent learning performance, and... | [
"point cloud",
"coordinate frame",
"robot manipulation",
"3D",
"RL"
] | By mining coordinate frames of input point clouds, we can significantly improve the training efficiency of 3D manipulation learning without changing camera placements or adding extra cameras. | 32 | 2210.07442 | title_snapshot |
_u4m5aEbWfU | SurroundDepth: Entangling Surrounding Views for Self-Supervised Multi-Camera Depth Estimation | https://openreview.net/forum?id=_u4m5aEbWfU | [
"Yi Wei",
"Linqing Zhao",
"Wenzhao Zheng",
"Zheng Zhu",
"Yongming Rao",
"Guan Huang",
"Jiwen Lu",
"Jie Zhou"
] | Poster | null | Depth estimation from images serves as the fundamental step of 3D perception for autonomous driving and is an economical alternative to expensive depth sensors like LiDAR. The temporal photometric consistency enables self-supervised depth estimation without labels, further facilitating its application. However, most ex... | [
"Self-supervised depth estimation",
"Multi-camera perception",
"Structure-from-motion"
] | null | 33 | 2204.03636 | title_snapshot |
ZUtgUA0Fuwd | Watch and Match: Supercharging Imitation with Regularized Optimal Transport | https://openreview.net/forum?id=ZUtgUA0Fuwd | [
"Siddhant Haldar",
"Vaibhav Mathur",
"Denis Yarats",
"Lerrel Pinto"
] | Oral | null | Imitation learning holds tremendous promise in learning policies efficiently for complex decision making problems. Current state-of-the-art algorithms often use inverse reinforcement learning (IRL), where given a set of expert demonstrations, an agent alternatively infers a reward function and the associated optimal po... | [
"Imitation Learning",
"Manipulation",
"Robotics"
] | We propose a new imitation learning algorithm that substantially improves sample efficiency for continuous control problems in simulation and on real-world robotic manipulation tasks. | 35 | 2206.15469 | title_snapshot |
uhIfIEIiWm_ | Offline Reinforcement Learning for Visual Navigation | https://openreview.net/forum?id=uhIfIEIiWm_ | [
"Dhruv Shah",
"Arjun Bhorkar",
"Hrishit Leen",
"Ilya Kostrikov",
"Nicholas Rhinehart",
"Sergey Levine"
] | Oral | null | Reinforcement learning can enable robots to navigate to distant goals while optimizing user-specified reward functions, including preferences for following lanes, staying on paved paths, or avoiding freshly mowed grass. However, online learning from trial-and-error for real-world robots is logistically challenging, and... | [
"offline reinforcement learning",
"visual navigation",
"motion planning"
] | Coupling the offline learned value function with a topological graph, where the values provide distance estimates, is a great way to make things scale. And it works with diverse rewards, like "stay in the sun", or "stay off my lawn"." | 41 | 2212.08244 | title_snapshot |
GeM6VUwYinO | One-Shot Transfer of Affordance Regions? AffCorrs! | https://openreview.net/forum?id=GeM6VUwYinO | [
"Denis Hadjivelichkov",
"Sicelukwanda Zwane",
"Lourdes Agapito",
"Marc Peter Deisenroth",
"Dimitrios Kanoulas"
] | Poster | null | In this work, we tackle one-shot visual search of object parts. Given a single reference image of an object with annotated affordance regions, we segment semantically corresponding parts within a target scene. We propose AffCorrs, an unsupervised model that combines the properties of pre-trained DINO-ViT's image desc... | [
"One-shot",
"Affordance",
"Correspondence"
] | One-shot transfer of affordance masks to unseen objects, both belonging to the same class and other classes. | 42 | 2209.07147 | title_snapshot |
gOW9pdOeqUc | Lidar Line Selection with Spatially-Aware Shapley Value for Cost-Efficient Depth Completion | https://openreview.net/forum?id=gOW9pdOeqUc | [
"Kamil Adamczewski",
"Christos Sakaridis",
"Vaishakh Patil",
"Luc Van Gool"
] | Poster | null | Lidar is a vital sensor for estimating the depth of a scene. Typical spinning lidars emit pulses arranged in several horizontal lines and the monetary cost of the sensor increases with the number of these lines. In this work, we present the new problem of optimizing the positioning of lidar lines to find the most effec... | [
"Lidar",
"Shapley value",
"feature selection",
"depth completion"
] | To reduce the cost of depth completion, we propose a spatially-aware Shapley value scheme for selecting the most important lines of a lidar | 43 | 2303.11720 | title_snapshot |
KDuBFXyRuE- | Iterative Interactive Modeling for Knotting Plastic Bags | https://openreview.net/forum?id=KDuBFXyRuE- | [
"Chongkai Gao",
"Zekun Li",
"Haichuan Gao",
"Feng Chen"
] | Poster | null | Deformable object manipulation has great research significance for the robotic community and numerous applications in daily life. In this work, we study how to knot plastic bags that are randomly dropped from the air with a dual-arm robot based on image input. The complex initial configuration and terrible material pro... | [
"Plastic Bag Manipulation",
"Learning from Demonstrations"
] | Knotting plastic bags randomly dropped from the air with a dual-arm robotic system and iterative interactive modeling. | 46 | null | null |
-8Kt5J-9AI | Tailoring Visual Object Representations to Human Requirements: A Case Study with a Recycling Robot | https://openreview.net/forum?id=-8Kt5J-9AI | [
"Debasmita Ghose",
"Michal Adam Lewkowicz",
"Kaleb Gezahegn",
"Julian Lee",
"Timothy Adamson",
"Marynel Vazquez",
"Brian Scassellati"
] | Poster | null | Robots are well-suited to alleviate the burden of repetitive and tedious manipulation tasks. In many applications though, a robot may be asked to interact with a wide variety of objects, making it hard or even impossible to pre-program visual object classifiers suitable for the task of interest. In this work, we study ... | [
"Representation Learning",
"Contrastive Learning",
"Human-in-the-loop Robot Learning"
] | null | 47 | null | null |
tJE1Yyi8fUX | DexPoint: Generalizable Point Cloud Reinforcement Learning for Sim-to-Real Dexterous Manipulation | https://openreview.net/forum?id=tJE1Yyi8fUX | [
"Yuzhe Qin",
"Binghao Huang",
"Zhao-Heng Yin",
"Hao Su",
"Xiaolong Wang"
] | Poster | null | We propose a sim-to-real framework for dexterous manipulation which can generalize to new objects of the same category in the real world. The key of our framework is to train the manipulation policy with point cloud inputs and dexterous hands. We propose two new techniques to enable joint learning on multiple objects a... | [
"Dexterous Manipulation",
"Policy Learning",
"Point Clouds",
"Sim-to-Real"
] | Learning generalizable dexterous manipulation with reinforcement learning using point cloud observation. | 50 | 2211.09423 | title_snapshot |
awciQcCEGJs | Graph Inverse Reinforcement Learning from Diverse Videos | https://openreview.net/forum?id=awciQcCEGJs | [
"Sateesh Kumar",
"Jonathan Zamora",
"Nicklas Hansen",
"Rishabh Jangir",
"Xiaolong Wang"
] | Oral | null | Research on Inverse Reinforcement Learning (IRL) from third-person videos has shown encouraging results on removing the need for manual reward design for robotic tasks. However, most prior works are still limited by training from a relatively restricted domain of videos. In this paper, we argue that the true potential ... | [
"Inverse Reinforcement Learning",
"Third-Person Video",
"Graph Network"
] | null | 53 | 2207.14299 | title_snapshot |
fnaMlJbRc4t | Interpretable Self-Aware Neural Networks for Robust Trajectory Prediction | https://openreview.net/forum?id=fnaMlJbRc4t | [
"Masha Itkina",
"Mykel Kochenderfer"
] | Poster | null | Although neural networks have seen tremendous success as predictive models in a variety of domains, they can be overly confident in their predictions on out-of-distribution (OOD) data. To be viable for safety-critical applications, like autonomous vehicles, neural networks must accurately estimate their epistemic or mo... | [
"autonomous vehicles",
"trajectory prediction",
"out-of-distribution detection",
"distribution shift",
"epistemic uncertainty estimation"
] | We propose the use of evidential deep learning to perform one-shot epistemic uncertainty estimation over a low-dimensional, interpretable latent space in a trajectory prediction setting. | 54 | 2211.08701 | title_snapshot |
h5g_VDJJfbt | Learning Generalizable Dexterous Manipulation from Human Grasp Affordance | https://openreview.net/forum?id=h5g_VDJJfbt | [
"Yueh-Hua Wu",
"Jiashun Wang",
"Xiaolong Wang"
] | Poster | null | Dexterous manipulation with a multi-finger hand is one of the most challenging problems in robotics. While recent progress in imitation learning has largely improved the sample efficiency compared to Reinforcement Learning, the learned policy can hardly generalize to manipulate novel objects, given limited expert demon... | [
"Generalized policy learning",
"Dexterous manipulation",
"Affordance model",
"Reinforcement learning",
"Imitation learning"
] | We propose a pipeline that generates large-scale demonstrations for dexterous manipulation and an imitation learning algorithm that generalizes to novel objects absent during training. | 55 | 2204.02320 | title_snapshot |
Mp3Y5jd7rnW | CADSim: Robust and Scalable in-the-wild 3D Reconstruction for Controllable Sensor Simulation | https://openreview.net/forum?id=Mp3Y5jd7rnW | [
"Jingkang Wang",
"Sivabalan Manivasagam",
"Yun Chen",
"Ze Yang",
"Ioan Andrei Bârsan",
"Anqi Joyce Yang",
"Wei-Chiu Ma",
"Raquel Urtasun"
] | Poster | null | Realistic simulation is key to enabling safe and scalable development of self-driving vehicles. A core component is simulating the sensors so that the entire autonomy system can be tested in simulation. Sensor simulation involves modeling traffic participants, such as vehicles, with high-quality appearance and articula... | [
"3D Reconstruction",
"CAD models",
"Sensor Simulation",
"Self-Driving"
] | We propose a new method to reconstruct objects from sensory observations that are of high fidelity, part-aware, geometry-aligned and compatible to graphics engine thus enable realistic and controllable simulation efficiently | 56 | 2311.01447 | title_snapshot |
HbGgF93Ppoy | Inferring Smooth Control: Monte Carlo Posterior Policy Iteration with Gaussian Processes | https://openreview.net/forum?id=HbGgF93Ppoy | [
"Joe Watson",
"Jan Peters"
] | Oral | null | Monte Carlo methods have become increasingly relevant for control of non-differentiable systems, approximate dynamics models, and learning from data.
These methods scale to high-dimensional spaces and are effective at the non-convex optimization often seen in robot learning. We look at sample-based methods from the per... | [
"model predictive control",
"policy search",
"reinforcement learning",
"approximate inference"
] | Sample-based MPC is a form for Bayesian inference. Use Gaussian processes for smooth actions and choose the temperature to have a healthy effective sample size. | 59 | 2210.03512 | title_snapshot |
lV-rNbXVSaO | Semantic Abstraction: Open-World 3D Scene Understanding from 2D Vision-Language Models | https://openreview.net/forum?id=lV-rNbXVSaO | [
"Huy Ha",
"Shuran Song"
] | Poster | null | We study open-world 3D scene understanding, a family of tasks that require agents to reason about their 3D environment with an open-set vocabulary and out-of-domain visual inputs -- a critical skill for robots to operate in the unstructured 3D world. Towards this end, we propose Semantic Abstraction (SemAbs), a framew... | [
"3D scene understanding",
"out-of-domain generalization",
"language"
] | We proposed Semantic Abstraction, a framework that equips 2D VLMs with 3D spatial capabilities for open-world 3D scene understanding tasks. | 61 | 2207.11514 | title_snapshot |
qUhkhHw8Dz | VideoDex: Learning Dexterity from Internet Videos | https://openreview.net/forum?id=qUhkhHw8Dz | [
"Kenneth Shaw",
"Shikhar Bahl",
"Deepak Pathak"
] | Poster | null | To build general robotic agents that can operate in many environments, it is often imperative for the robot to collect experience in the real world. However, this is often not feasible due to safety, time and hardware restrictions. We thus propose leveraging the next best thing as real world experience: internet vide... | [
"Dexterous Manipulation",
"Large Scale Robotics",
"Imitation Learning"
] | Learning Robot Action Priors from Human Videos | 63 | 2212.04498 | title_snapshot |
RgJwDQwW82y | Last-Mile Embodied Visual Navigation | https://openreview.net/forum?id=RgJwDQwW82y | [
"Justin Wasserman",
"Karmesh Yadav",
"Girish Chowdhary",
"Abhinav Gupta",
"Unnat Jain"
] | Poster | null | Realistic long-horizon tasks like image-goal navigation involve exploratory and exploitative phases. Assigned with an image of the goal, an embodied agent must explore to discover the goal, i.e., search efficiently using learned priors. Once the goal is discovered, the agent must accurately calibrate the last-mile of n... | [
"Visual Navigation",
"Embodied AI",
"Image-Goal Navigation",
"Perspective-n-Point",
"AI Habitat",
"Sim-to-Real"
] | A last-mile navigation module that connects to prior policies, leading to improved image-goal navigation results in simulation and real-robot experiments. | 64 | 2211.11746 | title_snapshot |
CC4JMO4dzg | Learning Preconditions of Hybrid Force-Velocity Controllers for Contact-Rich Manipulation | https://openreview.net/forum?id=CC4JMO4dzg | [
"Jacky Liang",
"Xianyi Cheng",
"Oliver Kroemer"
] | Poster | null | Robots need to manipulate objects in constrained environments like shelves and cabinets when assisting humans in everyday settings like homes and offices. These constraints make manipulation difficult by reducing grasp accessibility, so robots need to use non-prehensile strategies that leverage object-environment conta... | [
"Contact-Rich Manipulation",
"Hybrid Force-Velocity Controllers",
"Precondition Learning"
] | Plan with learned preconditions for Hybrid Force-Velocity Controllers to perform contact-rich manipulation tasks in constrained environments (e.g. shelves) | 65 | 2206.12728 | title_snapshot |
AdFROt9BoqE | Cross-Domain Transfer via Semantic Skill Imitation | https://openreview.net/forum?id=AdFROt9BoqE | [
"Karl Pertsch",
"Ruta Desai",
"Vikash Kumar",
"Franziska Meier",
"Joseph J Lim",
"Dhruv Batra",
"Akshara Rai"
] | Poster | null | We propose an approach for semantic imitation, which uses demonstrations from a source domain, e.g. human videos, to accelerate reinforcement learning (RL) in a different target domain, e.g. a robotic manipulator in a simulated kitchen. Instead of imitating low-level actions like joint velocities, our approach imitates... | [
"Reinforcement Learning",
"Imitation",
"Transfer Learning"
] | We propose an approach for semantic imitation, which uses demonstrations from a source domain, e.g. human videos, to accelerate reinforcement learning (RL) in a different target domain, e.g. a robotic manipulator in a simulated kitchen. | 67 | 2212.07407 | title_snapshot |
OIaJRUo5UXy | Learning Neuro-Symbolic Skills for Bilevel Planning | https://openreview.net/forum?id=OIaJRUo5UXy | [
"Tom Silver",
"Ashay Athalye",
"Joshua B. Tenenbaum",
"Tomás Lozano-Pérez",
"Leslie Pack Kaelbling"
] | Poster | null | Decision-making is challenging in robotics environments with continuous object-centric states, continuous actions, long horizons, and sparse feedback. Hierarchical approaches, such as task and motion planning (TAMP), address these challenges by decomposing decision-making into two or more levels of abstraction. In a se... | [
"Skill Learning",
"Neuro-Symbolic",
"Task and Motion Planning"
] | We learn neuro-symbolic skills from demonstrations and plan with them using search-then-sample TAMP techniques. | 68 | 2206.10680 | title_snapshot |
1zbWQxFIU- | MegaPose: 6D Pose Estimation of Novel Objects via Render & Compare | https://openreview.net/forum?id=1zbWQxFIU- | [
"Yann Labbé",
"Lucas Manuelli",
"Arsalan Mousavian",
"Stephen Tyree",
"Stan Birchfield",
"Jonathan Tremblay",
"Justin Carpentier",
"Mathieu Aubry",
"Dieter Fox",
"Josef Sivic"
] | Poster | null | We introduce MegaPose, a method to estimate the 6D pose of novel objects, that is, objects unseen during training. At inference time, the method only assumes knowledge of (i) a region of interest displaying the object in the image and (ii) a CAD model of the observed object. The contributions of this work are threefold... | [
"Object Pose Estimation",
"Robot Manipulation",
"Computer Vision"
] | A method for 6D pose estimation trained on synthetic data that can be used for any previously unseen object with CAD model. | 70 | 2212.06870 | title_snapshot |
qzMY915hCYX | Safety-Enhanced Autonomous Driving Using Interpretable Sensor Fusion Transformer | https://openreview.net/forum?id=qzMY915hCYX | [
"Hao Shao",
"Letian Wang",
"Ruobing Chen",
"Hongsheng Li",
"Yu Liu"
] | Poster | null | Large-scale deployment of autonomous vehicles has been continually delayed due to safety concerns. On the one hand, comprehensive scene understanding is indispensable, a lack of which would result in vulnerability to rare but complex traffic situations, such as the sudden emergence of unknown objects. However, reasonin... | [
"autonomous driving",
"sensor fusion",
"transformer",
"safety"
] | null | 71 | 2207.14024 | title_snapshot |
H6rr_CGzV9y | A Dual Representation Framework for Robot Learning with Human Guidance | https://openreview.net/forum?id=H6rr_CGzV9y | [
"Ruohan Zhang",
"Dhruva Bansal",
"Yilun Hao",
"Ayano Hiranaka",
"Jialu Gao",
"Chen Wang",
"Roberto Martín-Martín",
"Li Fei-Fei",
"Jiajun Wu"
] | Poster | null | The ability to interactively learn skills from human guidance and adjust behavior according to human preference is crucial to accelerating robot learning. But human guidance is an expensive resource, calling for methods that can learn efficiently. In this work, we argue that learning is more efficient if the agent is e... | [
"Human Guidance",
"Evaluative Feedback",
"Preference Learning"
] | A robotic agent learns more efficiently from human guidance with two representations, one for learning low-level control policies, the other for representing internal states of humans. | 72 | null | null |
UWTP_JvSug | Proactive slip control by learned slip model and trajectory adaptation | https://openreview.net/forum?id=UWTP_JvSug | [
"Kiyanoush Nazari",
"Willow Mandil",
"Amir Masoud Ghalamzan Esfahani"
] | Poster | null | This paper presents a novel control approach to dealing with a slip during robotic manipulative movements. Slip is a major cause of failure in many robotic grasping and manipulation tasks. Existing works use increased gripping forces to avoid/control slip. However, this may not be feasible, e.g., because (i) the robot ... | [
"Manipulation",
"Motion planning",
"Slip control",
"Motion control"
] | null | 75 | 2209.06019 | title_snapshot |
MUcBYHjzqp7 | Transferring Hierarchical Structures with Dual Meta Imitation Learning | https://openreview.net/forum?id=MUcBYHjzqp7 | [
"Chongkai Gao",
"Yizhou Jiang",
"Feng Chen"
] | Poster | null | Hierarchical Imitation Learning (HIL) is an effective way for robots to learn sub-skills from long-horizon unsegmented demonstrations. However, the learned hierarchical structure lacks the mechanism to transfer across multi-tasks or to new tasks, which makes them have to learn from scratch when facing a new situation. ... | [
"Hierarchical Imitation Learning",
"Meta Learning"
] | A dual meta imitation learning framework for fast adaptation of hierarchical imitation learning methods in new tasks. | 76 | 2201.11981 | title_judge |
tVgD4METs6o | Motion Style Transfer: Modular Low-Rank Adaptation for Deep Motion Forecasting | https://openreview.net/forum?id=tVgD4METs6o | [
"Parth Kothari",
"Danya Li",
"Yuejiang Liu",
"Alexandre Alahi"
] | Poster | null | Deep motion forecasting models have achieved great success when trained on a massive amount of data. Yet, they often perform poorly when training data is limited. To address this challenge, we propose a transfer learning approach for efficiently adapting pre-trained forecasting models to new domains, such as unseen age... | [
"Motion Forecasting",
"Trajectory Prediction",
"Distribution Shifts",
"Transfer Learning"
] | We propose a modular low-rank adaptation method that enables fast adaptation of deep motion forecasting models to new styles with limited samples | 78 | 2211.03165 | title_snapshot |
PS_eCS_WCvD | Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation | https://openreview.net/forum?id=PS_eCS_WCvD | [
"Mohit Shridhar",
"Lucas Manuelli",
"Dieter Fox"
] | Poster | null | Transformers have revolutionized vision and natural language processing with their ability to scale with large datasets. But in robotic manipulation, data is both limited and expensive. Can manipulation still benefit from Transformers with the right problem formulation? We investigate this question with PerAct, a langu... | [
"Transformers",
"Language Grounding",
"Manipulation",
"Behavior Cloning"
] | A transformer trained to 'detect actions' can imitate a wide variety of tabletop manipulation tasks. | 93 | 2209.05451 | title_snapshot |
WJbw_C-pCox | Synthesizing Adversarial Visual Scenarios for Model-Based Robotic Control | https://openreview.net/forum?id=WJbw_C-pCox | [
"Shubhankar Agarwal",
"Sandeep P. Chinchali"
] | Poster | null | Today’s robots often interface data-driven perception and planning models with classical model-predictive controllers (MPC). Often, such learned perception/planning models produce erroneous waypoint predictions on out-of-distribution (OoD) or even adversarial visual inputs, which increase control cost. However, today’s... | [] | null | 94 | 2204.06173 | title_snapshot |
wyCdmAJJY1F | CausalAF: Causal Autoregressive Flow for Safety-Critical Driving Scenario Generation | https://openreview.net/forum?id=wyCdmAJJY1F | [
"Wenhao Ding",
"Haohong Lin",
"Bo Li",
"Ding Zhao"
] | Poster | null | Generating safety-critical scenarios, which are crucial yet difficult to collect, provides an effective way to evaluate the robustness of autonomous driving systems. However, the diversity of scenarios and efficiency of generation methods are heavily restricted by the rareness and structure of safety-critical scenarios... | [
"Causal Generative Models",
"Scenario Generation",
"Autonomous Driving"
] | We propose a causal generative model for safety-critical driving scenarios generation to improve the robustness of autonomous driving algorithms. | 95 | 2110.13939 | title_snapshot |
_8DoIe8G3t | BEHAVIOR-1K: A Benchmark for Embodied AI with 1,000 Everyday Activities and Realistic Simulation | https://openreview.net/forum?id=_8DoIe8G3t | [
"Chengshu Li",
"Ruohan Zhang",
"Josiah Wong",
"Cem Gokmen",
"Sanjana Srivastava",
"Roberto Martín-Martín",
"Chen Wang",
"Gabrael Levine",
"Michael Lingelbach",
"Jiankai Sun",
"Mona Anvari",
"Minjune Hwang",
"Manasi Sharma",
"Arman Aydin",
"Dhruva Bansal",
"Samuel Hunter",
"Kyu-Young ... | Oral | null | We present BEHAVIOR-1K, a comprehensive simulation benchmark for human-centered robotics. BEHAVIOR-1K includes two components, guided and motivated by the results of an extensive survey on "what do you want robots to do for you?". The first is the definition of 1,000 everyday activities, grounded in 50 scenes (houses, ... | [
"Embodied AI Benchmark",
"Everyday Activities",
"Mobile Manipulation"
] | BEHAVIOR-1K is a novel human-centric benchmark for Embodied AI in simulation with 1000 everyday activities, a diverse dataset of 5,000+ objects and 50 scenes, and a simulation environment, OmniGibson, that reaches high levels of simulation realism. | 98 | null | null |
ndYsaoyzCWv | Temporal Logic Imitation: Learning Plan-Satisficing Motion Policies from Demonstrations | https://openreview.net/forum?id=ndYsaoyzCWv | [
"Yanwei Wang",
"Nadia Figueroa",
"Shen Li",
"Ankit Shah",
"Julie Shah"
] | Oral | null | Learning from demonstration (LfD) has successfully solved tasks featuring a long time horizon. However, when the problem complexity also includes human-in-the-loop perturbations, state-of-the-art approaches do not guarantee the successful reproduction of a task. In this work, we identify the roots of this challenge as ... | [
"Learning from Demonstration",
"Dynamical Systems",
"Formal Methods",
"Linear Temporal Logic",
"Certifiable Imitation Learning"
] | Combining the task-level reactivity of LTL and the motion-level reactivity of DS, we arrive at an imitation learning system able to robustly perform various multi-step tasks under arbitrary perturbations given only a small number of demonstrations. | 99 | 2206.04632 | title_snapshot |
SrSCqW4dq9 | Volumetric-based Contact Point Detection for 7-DoF Grasping | https://openreview.net/forum?id=SrSCqW4dq9 | [
"Junhao Cai",
"Jingcheng Su",
"Zida Zhou",
"Hui Cheng",
"Qifeng Chen",
"Michael Y Wang"
] | Poster | null | In this paper, we propose a novel grasp pipeline based on contact point detection on the truncated signed distance function (TSDF) volume to achieve closed-loop 7-degree-of-freedom (7-DoF) grasping on cluttered environments. The key aspects of our method are that 1) the proposed pipeline exploits the TSDF volume in ter... | [
"Contact point detection",
"7-DoF grasping",
"General object grasping"
] | null | 100 | 2209.06675 | title_snapshot |
esOrVR_8-rc | Generalization with Lossy Affordances: Leveraging Broad Offline Data for Learning Visuomotor Tasks | https://openreview.net/forum?id=esOrVR_8-rc | [
"Kuan Fang",
"Patrick Yin",
"Ashvin Nair",
"Homer Rich Walke",
"Gengchen Yan",
"Sergey Levine"
] | Oral | null | The use of broad datasets has proven to be crucial for generalization for a wide range of fields. However, how to effectively make use of diverse multi-task data for novel downstream tasks still remains a grand challenge in reinforcement learning and robotics. To tackle this challenge, we introduce a framework that acq... | [
"Reinforcement Learning",
"Representation Learning",
"Planning"
] | A framework that leverages diverse offline data for learning representations, goal-conditioned policies, and affordance models that enable rapid fine-tuning to new tasks in target scenes. | 104 | 2210.06601 | title_snapshot |
LunGpKUNIR | SE(3)-Equivariant Relational Rearrangement with Neural Descriptor Fields | https://openreview.net/forum?id=LunGpKUNIR | [
"Anthony Simeonov",
"Yilun Du",
"Yen-Chen Lin",
"Alberto Rodriguez Garcia",
"Leslie Pack Kaelbling",
"Tomás Lozano-Pérez",
"Pulkit Agrawal"
] | Poster | null | We present a framework for specifying tasks involving spatial relations between objects using only 5-10 demonstrations and then executing such tasks given point cloud observations of a novel pair of objects in arbitrary initial poses. Our approach structures these rearrangement tasks by assigning a consistent local coo... | [
"Neural Fields",
"Relations",
"Manipulation",
"Object Rearrangement"
] | We present a method to execute relational rearrangement tasks on pairs of unseen objects using a few demonstrations. | 105 | 2211.09786 | title_snapshot |
ZL2keFk7WXJ | Learning Multi-Objective Curricula for Robotic Policy Learning | https://openreview.net/forum?id=ZL2keFk7WXJ | [
"Jikun Kang",
"Miao Liu",
"Abhinav Gupta",
"Christopher Pal",
"Xue Liu",
"Jie Fu"
] | Poster | null | Various automatic curriculum learning (ACL) methods have been proposed to improve the sample efficiency and final performance of robots' policies learning. They are designed to control how a robotic agent collects data, which is inspired by how humans gradually adapt their learning processes to their capabilities. In t... | [
"ACL",
"Hyper-net",
"Multi-objective Curricula"
] | null | 107 | 2110.03032 | title_snapshot |
r-w9Wh-QVnH | Real-time Mapping of Physical Scene Properties with an Autonomous Robot Experimenter | https://openreview.net/forum?id=r-w9Wh-QVnH | [
"Iain Haughton",
"Edgar Sucar",
"Andre Mouton",
"Edward Johns",
"Andrew Davison"
] | Oral | null | Neural fields can be trained from scratch to represent the shape and appearance of 3D scenes efficiently. It has also been shown that they can densely map correlated properties such as semantics, via sparse interactions from a human labeller. In this work, we show that a robot can densely annotate a scene with arbitrar... | [
"Neural field",
"robot experimentation",
"physical properties"
] | An autonomous robot experimenter discovers dense physical scene properties by providing sparse interaction measurements to a 3D neural field. | 108 | 2210.17325 | title_snapshot |
BxHcg_Zlpxj | Rethinking Sim2Real: Lower Fidelity Simulation Leads to Higher Sim2Real Transfer in Navigation | https://openreview.net/forum?id=BxHcg_Zlpxj | [
"Joanne Truong",
"Max Rudolph",
"Naoki Harrison Yokoyama",
"Sonia Chernova",
"Dhruv Batra",
"Akshara Rai"
] | Poster | null | If we want to train robots in simulation before deploying them in reality, it seems natural and almost self-evident to presume that reducing the sim2real gap involves creating simulators of increasing fidelity (since reality is what it is). We challenge this assumption and present a contrary hypothesis -- sim2real tran... | [
"Sim2Real",
"Deep Reinforcement Learning",
"Visual-Based Navigation"
] | Sim2real transfer of robots may be improved not by increasing but by decreasing simulation fidelity; we should instead prioritize simulation speed for tasks that can be represented with abstract action spaces. | 112 | 2207.10821 | title_snapshot |
b1zMsZuzd2e | Learning Dense Visual Descriptors using Image Augmentations for Robot Manipulation Tasks | https://openreview.net/forum?id=b1zMsZuzd2e | [
"Christian Graf",
"David B. Adrian",
"Joshua Weil",
"Miroslav Gabriel",
"Philipp Schillinger",
"Markus Spies",
"Heiko Neumann",
"Andras Gabor Kupcsik"
] | Poster | null | We propose a self-supervised training approach for learning view-invariant
dense visual descriptors using image augmentations. Unlike existing
works, which often require complex datasets, such as registered RGBD sequences,
we train on an unordered set of RGB images. This allows for learning from a single
camera view, ... | [
"self-supervised learning",
"computer vision",
"representation learning",
"bin-picking"
] | We propose a self-supervised training approach for learning view-invariant dense visual descriptors using image augmentations. | 114 | 2209.05213 | title_snapshot |
th7GW868Pok | Proactive Robot Assistance via Spatio-Temporal Object Modeling | https://openreview.net/forum?id=th7GW868Pok | [
"Maithili Patel",
"Sonia Chernova"
] | Poster | null | Proactive robot assistance enables a robot to anticipate and provide for a user's needs without being explicitly asked. We formulate proactive assistance as the problem of the robot anticipating temporal patterns of object movements associated with everyday user routines, and proactively assisting the user by placing o... | [
"Proactive Robot Assistance",
"Spatio-Temporal Object Tracking"
] | A graph translation network based model that can learn to predict robot actions by modeling future movement of objects in an environment resulting from the user’s routine activities, enabling a robot to assist proactively, without being asked. | 119 | 2211.15501 | title_snapshot |
tGbpgz6yOrI | R3M: A Universal Visual Representation for Robot Manipulation | https://openreview.net/forum?id=tGbpgz6yOrI | [
"Suraj Nair",
"Aravind Rajeswaran",
"Vikash Kumar",
"Chelsea Finn",
"Abhinav Gupta"
] | Poster | null | We study how visual representations pre-trained on diverse human video data can enable data-efficient learning of downstream robotic manipulation tasks. Concretely, we pre-train a visual representation using the Ego4D human video dataset using a combination of time-contrastive learning, video-language alignment, and an... | [
"Visual Representation Learning",
"Robotic Manipulation"
] | Pre-training a visual representation on diverse human video datasets, that can be downloaded and used off-the-shelf to enable more data efficient robot learning in simulation and the real world. | 121 | 2203.12601 | title_snapshot |
PZiKO7mjC43 | Towards Capturing the Temporal Dynamics for Trajectory Prediction: a Coarse-to-Fine Approach | https://openreview.net/forum?id=PZiKO7mjC43 | [
"Xiaosong Jia",
"Li Chen",
"Penghao Wu",
"Jia Zeng",
"Junchi Yan",
"Hongyang Li",
"Yu Qiao"
] | Poster | null | Trajectory prediction is one of the basic tasks in the autonomous driving field, which aims to predict the future position of other agents around the ego vehicle so that a safe yet efficient driving plan could be generated in the downstream module. Recently, deep learning based methods dominate the field. State-of-the... | [
"autonomous prediction",
"trajectory prediction"
] | We explore ways of modeling the temporal corelation of predicted time-steps in the trajectory prediction task. | 124 | null | null |
uhhA2OryTjj | Robust Trajectory Prediction against Adversarial Attacks | https://openreview.net/forum?id=uhhA2OryTjj | [
"Yulong Cao",
"Danfei Xu",
"Xinshuo Weng",
"Zhuoqing Mao",
"Anima Anandkumar",
"Chaowei Xiao",
"Marco Pavone"
] | Oral | null | Trajectory prediction using deep neural networks (DNNs) is an essential component of autonomous driving (AD) systems. However, these methods are vulnerable to adversarial attacks, leading to serious consequences such as collisions. In this work, we identify two key ingredients to defend trajectory prediction models ag... | [
"Adversarial Attack",
"Trajectory Prediction",
"Autonomous Driving"
] | We propose a new adversarial training framework for training robust trajectory prediction systems by addressing domain specific challenges. | 125 | 2208.00094 | title_snapshot |
zldI4UpuG7v | Deep Whole-Body Control: Learning a Unified Policy for Manipulation and Locomotion | https://openreview.net/forum?id=zldI4UpuG7v | [
"Zipeng Fu",
"Xuxin Cheng",
"Deepak Pathak"
] | Oral | null | An attached arm can significantly increase the applicability of legged robots to several mobile manipulation tasks that are not possible for the wheeled or tracked counterparts. The standard modular control pipeline for such legged manipulators is to decouple the controller into that of manipulation and locomotion. How... | [
"Mobile Manipulation",
"Whole-Body Control",
"Legged Locomotion"
] | Learning a unified policy for whole-body control of both the arm and legs of a custom-built low-cost quadruped mobile manipulator | 127 | 2210.10044 | title_snapshot |
7ZcePvChS7u | Human-Robot Commensality: Bite Timing Prediction for Robot-Assisted Feeding in Groups | https://openreview.net/forum?id=7ZcePvChS7u | [
"Jan Ondras",
"Abrar Anwar",
"Tong Wu",
"Fanjun Bu",
"Malte Jung",
"Jorge Jose Ortiz",
"Tapomayukh Bhattacharjee"
] | Poster | null | We develop data-driven models to predict when a robot should feed during social dining scenarios. Being able to eat independently with friends and family is considered one of the most memorable and important activities for people with mobility limitations. While existing robotic systems for feeding people with mobility... | [
"Multimodal Learning",
"HRI",
"Assistive Robotics",
"Group Dynamics"
] | We use our self-collected dataset to develop data-driven models that predicts when a robot should feed during social dining scenarios and evaluate with user study. | 128 | 2207.03348 | title_snapshot |
xK-UtqDpD7L | Learning to Grasp the Ungraspable with Emergent Extrinsic Dexterity | https://openreview.net/forum?id=xK-UtqDpD7L | [
"Wenxuan Zhou",
"David Held"
] | Oral | null | A simple gripper can solve more complex manipulation tasks if it can utilize the external environment such as pushing the object against the table or a vertical wall, known as "Extrinsic Dexterity." Previous work in extrinsic dexterity usually has careful assumptions about contacts which impose restrictions on robot de... | [
"Manipulation",
"Reinforcement Learning",
"Extrinsic Dexterity"
] | We build a system for the "Occluded Grasping" task with the combination of reinforcement learning and extrinsic dexterity that works on a real robot. | 132 | 2211.01500 | title_snapshot |
xC-68ANJeK_ | Learning Control Admissibility Models with Graph Neural Networks for Multi-Agent Navigation | https://openreview.net/forum?id=xC-68ANJeK_ | [
"Chenning Yu",
"Hongzhan Yu",
"Sicun Gao"
] | Poster | null | Deep reinforcement learning in continuous domains focuses on learning control policies that map states to distributions over actions that ideally concentrate on the optimal choices in each step. In multi-agent navigation problems, the optimal actions depend heavily on the agents' density. Their interaction patterns gro... | [
"Graph Neural Networks",
"Multi-Agent Navigation"
] | Control Admissibility Model decouples the properties of goal-reaching and collision avoidance, and is compositional which enables generalization to challengingly high density of agents. | 133 | 2210.09378 | title_snapshot |
KXkzplx6H2K | Socially-Attentive Policy Optimization in Multi-Agent Self-Driving System | https://openreview.net/forum?id=KXkzplx6H2K | [
"Zipeng Dai",
"Tianze Zhou",
"Kun Shao",
"David Henry Mguni",
"Bin Wang",
"Jianye HAO"
] | Poster | null | As increasing numbers of autonomous vehicles (AVs) are being deployed, it is important to construct a multi-agent self-driving (MASD) system for navigating traffic flows of AVs. In an MASD system, AVs not only navigate themselves to pursue their own goals, but also interact with each other to prevent congestion or coll... | [
"self-driving",
"social preference",
"multi-agent",
"reinforcement learning"
] | We focus on how to improve the safety and efficiency of a multi-agent self-driving system, by navigating vehicles to learn socially-compatible behaviors. | 134 | null | null |
f_XmiyZcsjL | Reciprocal MIND MELD: Improving Learning From Demonstration via Personalized, Reciprocal Teaching | https://openreview.net/forum?id=f_XmiyZcsjL | [
"Mariah L Schrum",
"Erin Hedlund-Botti",
"Matthew Gombolay"
] | Poster | null | Endowing robots with the ability to learn novel tasks via demonstrations will increase the accessibility of robots for non-expert, non-roboticists. However, research has shown that humans can be poor teachers, making it difficult for robots to effectively learn from humans. If the robot could instruct humans how to pro... | [
"meta-learning",
"personalization",
"imitation learning"
] | To reduce the problem of suboptimal demonstrations, we introduce Reciprocal MIND MELD which learns a semantically meaningful embedding space to provide actionable robotic feedback to demonstrators to improve upon the quality of their demonstrations. | 136 | null | null |
aQnn9cIVTRJ | Motion Policy Networks | https://openreview.net/forum?id=aQnn9cIVTRJ | [
"Adam Fishman",
"Adithyavairavan Murali",
"Clemens Eppner",
"Bryan Peele",
"Byron Boots",
"Dieter Fox"
] | Poster | null | Collision-free motion generation in unknown environments is a core building block for robot manipulation.
Generating such motions is challenging due to multiple objectives; not only should the solutions be optimal, the motion generator itself must be fast enough for real-time performance and reliable enough for practi... | [
"Motion Control",
"Imitation Learning",
"End-to-End Learning"
] | Motion Policy Networks are trained on millions of example trajectories to generate collision-free, smooth motion from just a single depth camera image | 138 | 2210.12209 | title_snapshot |
iM932PeUi_7 | Decentralized Data Collection for Robotic Fleet Learning: A Game-Theoretic Approach | https://openreview.net/forum?id=iM932PeUi_7 | [
"Oguzhan Akcin",
"Po-han Li",
"Shubhankar Agarwal",
"Sandeep P. Chinchali"
] | Poster | null | Fleets of networked autonomous vehicles (AVs) collect terabytes of sensory data, which is often transmitted to central servers (the ``cloud'') for training machine learning (ML) models. Ideally, these fleets should upload all their data, especially from rare operating contexts, in order to train robust ML models. Howev... | [] | null | 141 | null | null |
PAFEQQtDf8s | CoBEVT: Cooperative Bird’s Eye View Semantic Segmentation with Sparse Transformers | https://openreview.net/forum?id=PAFEQQtDf8s | [
"Runsheng Xu",
"Zhengzhong Tu",
"Hao Xiang",
"Wei Shao",
"Bolei Zhou",
"Jiaqi Ma"
] | Poster | null | Bird’s eye view (BEV) semantic segmentation plays a crucial role in spatial sensing for autonomous driving. Although recent literature has made significant progress on BEV map understanding, they are all based on single-agent camera-based systems. These solutions sometimes have difficulty handling occlusions or detecti... | [
"Autonomous driving",
"BEV map understanding",
"Vehicle-to-Vehicle (V2V) application"
] | We propose CoBEVT, the first generic multi-agent multi-camera perception framework that can cooperatively generate BEV map predictions via sparse Transformers. | 142 | 2207.02202 | title_snapshot |
2cEjfernc5P | Selective Object Rearrangement in Clutter | https://openreview.net/forum?id=2cEjfernc5P | [
"Bingjie Tang",
"Gaurav S. Sukhatme"
] | Poster | null | We propose an image-based, learned method for selective tabletop object rearrangement in clutter using a parallel jaw gripper. Our method consists of three stages: graph-based object sequencing (which object to move), feature-based action selection (whether to push or grasp, and at what position and orientation) and a ... | [
"Rearrangement",
"Robot Manipulation",
"Task and Motion Planning"
] | We propose a method for tabletop object rearrangement that deals with cluttered initial scene, target object selectivity, and occupied goal positions -- the first system to address all three concurrently in a purely image-based setting. | 147 | null | null |
Eal_lL08v_l | Transformers Are Adaptable Task Planners | https://openreview.net/forum?id=Eal_lL08v_l | [
"Vidhi Jain",
"Yixin Lin",
"Eric Undersander",
"Yonatan Bisk",
"Akshara Rai"
] | Poster | null | Every home is different, and every person likes things done in their particular way. Therefore, home robots of the future need to both reason about the sequential nature of day-to-day tasks and generalize to user's preferences. To this end, we propose a Transformer Task Planner (TTP) that learns high-level actions from... | [
"Task Planning",
"Prompt",
"Preferences",
"Object-centric Representation"
] | Transformer Task Planner (TTP) learns high-level pick and place actions from dishwasher loading demonstrations and adapts to unseen preferences using single prompt. | 148 | 2207.02442 | title_snapshot |
2gfB_kMVFvP | ToolFlowNet: Robotic Manipulation with Tools via Predicting Tool Flow from Point Clouds | https://openreview.net/forum?id=2gfB_kMVFvP | [
"Daniel Seita",
"Yufei Wang",
"Sarthak J Shetty",
"Edward Yao Li",
"Zackory Erickson",
"David Held"
] | Poster | null | Point clouds are a widely available and canonical data modality which convey the 3D geometry of a scene. Despite significant progress in classification and segmentation from point clouds, policy learning from such a modality remains challenging, and most prior works in imitation learning focus on learning policies from... | [
"Flow",
"Point Clouds",
"Tool Manipulation",
"Deformables"
] | A neural network which given a segmented point cloud, predicts per-point flow on a tool for manipulation. | 155 | 2211.09006 | title_snapshot |
y1t0c5eUk2w | Fusing Priori and Posteriori Metrics for Automatic Dataset Annotation of Planar Grasping | https://openreview.net/forum?id=y1t0c5eUk2w | [
"Hao Sha",
"Lai Qianen",
"Hongxiang Yu",
"Rong Xiong",
"Yue Wang"
] | Poster | null | Grasp detection based on deep learning has been a research hot spot in recent years. The performance of grasping detection models relies on high-quality, large-scale grasp datasets. Taking comprehensive consideration of quality, extendability, and annotation cost, metric-based simulation methodology is the most promis... | [
"Grasp Detection",
"Grasp Metric",
"Grasp Dataset",
"Automatic Annotation"
] | null | 157 | null | null |
ZAbh4Zp0Blt | PoET: Pose Estimation Transformer for Single-View, Multi-Object 6D Pose Estimation | https://openreview.net/forum?id=ZAbh4Zp0Blt | [
"Thomas Georg Jantos",
"Mohamed Amin Hamdad",
"Wolfgang Granig",
"Stephan Weiss",
"Jan Steinbrener"
] | Poster | null | Accurate 6D object pose estimation is an important task for a variety of robotic applications such as grasping or localization. It is a challenging task due to object symmetries, clutter and occlusion, but it becomes more challenging when additional information, such as depth and 3D models, is not provided. We present ... | [
"6D Pose Estimation",
"Transformer",
"Object-Relative Localization"
] | Novel transformer-based network for RGB-only, 6 degree of freedom multi-object pose estimation while not relying on depth maps or 3D object models | 163 | 2211.14125 | title_snapshot |
X6CjiTWVRVr | Out-of-Dynamics Imitation Learning from Multimodal Demonstrations | https://openreview.net/forum?id=X6CjiTWVRVr | [
"Yiwen Qiu",
"Jialong Wu",
"Zhangjie Cao",
"Mingsheng Long"
] | Poster | null | Existing imitation learning works mainly assume that the demonstrator who collects demonstrations shares the same dynamics as the imitator. However, the assumption limits the usage of imitation learning, especially when collecting demonstrations for the imitator is difficult. In this paper, we study out-of-dynamics imi... | [
"Imitation Learning",
"Out-of-Dynamics Imitation Learning"
] | We develop a transferability measurement learning by a sequence-based contrastive clustering algorithm and a adversarial-learning based algorithm to address out-of-dynamics imitation learning. | 175 | 2211.06839 | title_snapshot |
Z8mqodASTQd | Domain Adaptation and Generalization: A Low-Complexity Approach | https://openreview.net/forum?id=Z8mqodASTQd | [
"Joshua Niemeijer",
"Jörg Peter Schäfer"
] | Poster | null | Well-performing deep learning methods are essential in today's perception of robotic systems such as autonomous driving vehicles.
Ongoing research is due to the real-life demands for robust deep learning models against numerous domain changes and cheap training processes to avoid costly manual-labeling efforts.
These r... | [
"unsupervised domain adaptation",
"semantic segmentation",
"domain generalization"
] | null | 178 | null | null |
niys1Zt1blq | COACH: Cooperative Robot Teaching | https://openreview.net/forum?id=niys1Zt1blq | [
"Cunjun Yu",
"Yiqing Xu",
"Linfeng Li",
"David Hsu"
] | Poster | null | Knowledge and skills can transfer from human teachers to human students. However, such direct transfer is often not scalable for physical tasks, as they require one-to-one interaction, and human teachers are not available in sufficient numbers. Machine learning enables robots to become experts and play the role of tea... | [
"Robot Teaching",
"Human-Robot Interaction"
] | We present a conceptual framework, Cooperative Robot Teaching, which offers a formal model to describe and analyze robot teaching in a cooperative task. | 179 | 2302.06199 | title_snapshot |
qr0wqg8NqkL | TrackletMapper: Ground Surface Segmentation and Mapping from Traffic Participant Trajectories | https://openreview.net/forum?id=qr0wqg8NqkL | [
"Jannik Zürn",
"Sebastian Weber",
"Wolfram Burgard"
] | Poster | null | Robustly classifying ground infrastructure such as roads and street crossings is an essential task for mobile robots operating alongside pedestrians. While many semantic segmentation datasets are available for autonomous vehicles, models trained on such datasets exhibit a large domain gap when deployed on robots operat... | [
"Knowledge Distillation",
"Semantic Segmentation",
"Navigation"
] | We introduce TrackletMapper, an automatic image annotation method that leverages tracklets of traffic participants to semantically annotate images. | 181 | 2209.05247 | title_snapshot |
jTh3rdEF3LH | HUM3DIL: Semi-supervised Multi-modal 3D HumanPose Estimation for Autonomous Driving | https://openreview.net/forum?id=jTh3rdEF3LH | [
"Andrei Zanfir",
"Mihai Zanfir",
"Alex Gorban",
"Jingwei Ji",
"Yin Zhou",
"Dragomir Anguelov",
"Cristian Sminchisescu"
] | Poster | null | Autonomous driving is an exciting new industry, posing important research questions. Within the perception module, 3D human pose estimation is an emerging technology, which can enable the autonomous vehicle to perceive and understand the subtle and complex behaviors of pedestrians. While hardware systems and sensors ha... | [
"autonomous driving",
"perception",
"human pose",
"key points",
"skeletal representation"
] | 3D Human Pose estimation from RGB + LiDAR data, for autonomous driving. | 189 | 2212.07729 | title_judge |
u4QXJbcvx8u | Laplace Approximation Based Epistemic Uncertainty Estimation in 3D Object Detection | https://openreview.net/forum?id=u4QXJbcvx8u | [
"Peng Yun",
"Ming Liu"
] | Poster | null | Understanding the uncertainty of predictions is a desirable feature for perceptual modules in critical robotic applications. 3D object detectors are neural networks with high-dimensional output space. It suffers from poor calibration in classification and lacks reliable uncertainty estimation in regression. To provide ... | [
"Laplace approximation",
"epistemic uncertainty",
"3D object detection"
] | In this work, we tailor Laplace approximation for 3D object detection, and propose solutions in Fisher approximation, Bayesian inference, and weight prior determination. | 193 | null | null |
1Pt2FM5BLdi | Towards Online 3D Bin Packing: Learning Synergies between Packing and Unpacking via DRL | https://openreview.net/forum?id=1Pt2FM5BLdi | [
"Shuai Song",
"Shuo Yang",
"Ran Song",
"Shilei Chu",
"yibin Li",
"Wei Zhang"
] | Poster | null | There is an emerging research interest in addressing the online 3D bin packing problem (3D-BPP), which has a wide range of applications in logistics industry. However, neither heuristic methods nor those based on deep reinforcement learning (DRL) outperform human packers in real logistics scenarios. One important reaso... | [
"Bin packing",
"Robotics",
"Deep reinforcement learning"
] | null | 196 | null | null |
X_qYPtJLaX8 | PRISM: Probabilistic Real-Time Inference in Spatial World Models | https://openreview.net/forum?id=X_qYPtJLaX8 | [
"Atanas Mirchev",
"Baris Kayalibay",
"Ahmed Agha",
"Patrick van der Smagt",
"Daniel Cremers",
"Justin Bayer"
] | Oral | null | We introduce PRISM, a method for real-time filtering in a probabilistic generative model of agent motion and visual perception. Previous approaches either lack uncertainty estimates for the map and agent state, do not run in real-time, do not have a dense scene representation or do not model agent dynamics. Our solutio... | [
"generative model",
"SLAM",
"Bayes filter",
"uncertainty",
"differentiable rendering"
] | We propose a probabilistic real-time filtering inference for a generative spatial model that features rendering and dynamics. | 200 | 2212.02988 | title_snapshot |
NanNxv92Uih | Reinforcement learning with Demonstrations from Mismatched Task under Sparse Reward | https://openreview.net/forum?id=NanNxv92Uih | [
"Yanjiang Guo",
"Jingyue Gao",
"Zheng Wu",
"Chengming Shi",
"Jianyu Chen"
] | Poster | null | Reinforcement learning often suffer from the sparse reward issue in real-world robotics problems. Learning from demonstration (LfD) is an effective way to eliminate this problem, which leverages collected expert data to aid online learning. Prior works often assume that the learning agent and the expert aim to accompli... | [
"Sparse Reward Reinforcement Learning",
"Learn from Demonstration",
"Task Mismatch"
] | We propose CRSfD method to aid online reinforcement learning with demonstrations from mismatched task under sparse reward environment. | 204 | 2212.01509 | title_snapshot |
rbIzq-I84i_ | Graph network simulators can learn discontinuous, rigid contact dynamics | https://openreview.net/forum?id=rbIzq-I84i_ | [
"Kelsey R Allen",
"Tatiana Lopez Guevara",
"Yulia Rubanova",
"Kim Stachenfeld",
"Alvaro Sanchez-Gonzalez",
"Peter Battaglia",
"Tobias Pfaff"
] | Poster | null | Recent years have seen a rise in techniques for modeling discontinuous dynamics, such as rigid contact or switching motion modes, using deep learning. A common claim is that deep networks are incapable of accurately modeling rigid-body dynamics without explicit modules for handling contacts, due to the continuous natur... | [
"graph networks",
"contacts",
"rigid body dynamics",
"simulation"
] | Graph network simulators can accurately and efficiently model discontinuous, rigid body contact | 206 | null | null |
h0Yb0U_-Tki | Instruction-driven history-aware policies for robotic manipulations | https://openreview.net/forum?id=h0Yb0U_-Tki | [
"Pierre-Louis Guhur",
"Shizhe Chen",
"Ricardo Garcia Pinel",
"Makarand Tapaswi",
"Ivan Laptev",
"Cordelia Schmid"
] | Oral | null | In human environments, robots are expected to accomplish a variety of manipulation tasks given simple natural language instructions. Yet, robotic manipulation is extremely challenging as it requires fine-grained motor control, long-term memory as well as generalization to previously unseen tasks and environments. To ad... | [
"Robotics Manipulation",
"Language Instruction",
"Transformer"
] | A new model to solve robotics tasks using instructions, multi-views and history | 208 | 2209.04899 | title_snapshot |
7JVNhaMbZUu | Particle-Based Score Estimation for State Space Model Learning in Autonomous Driving | https://openreview.net/forum?id=7JVNhaMbZUu | [
"Angad Singh",
"Omar Makhlouf",
"Maximilian Igl",
"Joao Messias",
"Arnaud Doucet",
"Shimon Whiteson"
] | Poster | null | Multi-object state estimation is a fundamental problem for robotic applications where a robot must interact with other moving objects. Typically, other objects' relevant state features are not directly observable, and must instead be inferred from observations. Particle filtering can perform such inference given approx... | [
"Autonomous Driving",
"Particle Filtering",
"Self-supervised Learning"
] | Learning state space models for multi-object tracking in autonomous driving using particle methods. | 209 | 2212.06968 | title_snapshot |
Xo3eOibXCQ8 | Learning with Muscles: Benefits for Data-Efficiency and Robustness in Anthropomorphic Tasks | https://openreview.net/forum?id=Xo3eOibXCQ8 | [
"Isabell Wochner",
"Pierre Schumacher",
"Georg Martius",
"Dieter Büchler",
"Syn Schmitt",
"Daniel Haeufle"
] | Poster | null | Humans are able to outperform robots in terms of robustness, versatility, and learning of new tasks in a wide variety of movements. We hypothesize that highly nonlinear muscle dynamics play a large role in providing inherent stability, which is favorable to learning. While recent advances have been made in applying mod... | [
"reinforcement learning",
"model predictive control",
"actuator morphology",
"morphological computation"
] | We showcase the benefits of muscle-actuation for learning efficiency and robustness in a multitude of anthropomorphic tasks. | 210 | 2207.03952 | title_snapshot |
KwdUWypQ5gC | Bayesian Reinforcement Learning for Single-Episode Missions in Partially Unknown Environments | https://openreview.net/forum?id=KwdUWypQ5gC | [
"Matthew Budd",
"Paul Duckworth",
"Nick Hawes",
"Bruno Lacerda"
] | Poster | null | We consider planning for mobile robots conducting missions in real-world domains where a priori unknown dynamics affect the robot’s costs and transitions. We study single-episode missions where it is crucial that the robot appropriately trades off exploration and exploitation, such that the learning of the environment ... | [
"Planning under Uncertainty",
"Gaussian Processes",
"Single-Episode Bayesian Reinforcement Learning"
] | We propose a unified Bayes-optimal framework for single-mission robot planning in unknown environments, and evaluate on two realistic simulated environments. | 212 | null | null |
L8hCfhPbFho | VIOLA: Object-Centric Imitation Learning for Vision-Based Robot Manipulation | https://openreview.net/forum?id=L8hCfhPbFho | [
"Yifeng Zhu",
"Abhishek Joshi",
"Peter Stone",
"Yuke Zhu"
] | Poster | null | We introduce VIOLA, an object-centric imitation learning approach to learning closed-loop visuomotor policies for robot manipulation. Our approach constructs object-centric representations based on general object proposals from a pre-trained vision model. VIOLA uses a transformer-based policy to reason over these repre... | [
"Imitation Learning",
"Robot Manipulation",
"Object-Centric Representation"
] | We introduce an object-centric imitation learning approach for robot manipulation that acquire robust, closed-loop visuomotor poicy. | 217 | 2210.11339 | title_judge |
cF1dxVGxic- | Embedding Synthetic Off-Policy Experience for Autonomous Driving via Zero-Shot Curricula | https://openreview.net/forum?id=cF1dxVGxic- | [
"Eli Bronstein",
"Sirish Srinivasan",
"Supratik Paul",
"Aman Sinha",
"Matthew O'Kelly",
"Payam Nikdel",
"Shimon Whiteson"
] | Oral | null | ML-based motion planning is a promising approach to produce agents that exhibit complex behaviors, and automatically adapt to novel environments. In the context of autonomous driving, it is common to treat all available training data equally. However, this approach produces agents that do not perform robustly in safety... | [
"Imitation Learning",
"Curriculum Learning",
"Autonomous Driving"
] | A zero-shot transfer of a difficulty model to generate training curricula for an AV planning agent using only 10% of the data outperforms training on full dataset. | 219 | 2212.01375 | title_snapshot |
o8dLx8OVcNk | Learning Riemannian Stable Dynamical Systems via Diffeomorphisms | https://openreview.net/forum?id=o8dLx8OVcNk | [
"Jiechao Zhang",
"Hadi Beik Mohammadi",
"Leonel Rozo"
] | Poster | null | Dexterous and autonomous robots should be capable of executing elaborated dynamical motions skillfully. Learning techniques may be leveraged to build models of such dynamic skills. To accomplish this, the learning model needs to encode a stable vector field that resembles the desired motion dynamics. This is challengin... | [
"Dynamical systems",
"Riemannian manifolds",
"Motion learning",
"Learning from demonstrations",
"Neural ODEs"
] | Learning Lyapunov-stable Dynamical Systems on Riemannian Manifolds via Diffeormorphisms built on neural manifold ODEs. | 220 | 2211.03169 | title_snapshot |
bUUf1CT1sNu | Learning Robust Real-World Dexterous Grasping Policies via Implicit Shape Augmentation | https://openreview.net/forum?id=bUUf1CT1sNu | [
"Qiuyu Chen",
"Karl Van Wyk",
"Yu-Wei Chao",
"Wei Yang",
"Arsalan Mousavian",
"Abhishek Gupta",
"Dieter Fox"
] | Poster | null | Dexterous robotic hands have the capability to interact with a wide variety of household objects. However, learning robust real world grasping policies for arbitrary objects has proven challenging due to the difficulty of generating high quality training data. In this work, we propose a learning system (\emph{ISAGrasp}... | [
"Dexterous Manipulation",
"Learning from Human Demonstration",
"Data Augmentation"
] | We propose Implicit Shape Augmentation, a method that is able to robustly interact with daily unseen objects by leveraging simulation as well as a small number of human demonstrations. | 223 | 2210.13638 | title_snapshot |
K0gW0A6gi7G | Learning Visualization Policies of Augmented Reality for Human-Robot Collaboration | https://openreview.net/forum?id=K0gW0A6gi7G | [
"Kishan Dhananjay Chandan",
"Jack Albertson",
"Shiqi Zhang"
] | Poster | null | In human-robot collaboration domains, augmented reality (AR) technologies have enabled people to visualize the state of robots. Current AR-based visualization policies are designed manually, which requires a lot of human efforts and domain knowledge. When too little information is visualized, human users find the AR in... | [
"Augmented Reality",
"Multi-robot systems",
"Imitation Learning"
] | We present our framework, called VARIL, that for the first time introduces a learning-based Augmented Reality (AR) visualization strategy for human-multi-robot collaboration. | 224 | 2211.07028 | title_snapshot |
gJhuiYQ6VGJ | Deep Black-Box Reinforcement Learning with Movement Primitives | https://openreview.net/forum?id=gJhuiYQ6VGJ | [
"Fabian Otto",
"Onur Celik",
"Hongyi Zhou",
"Hanna Ziesche",
"Vien Anh Ngo",
"Gerhard Neumann"
] | Poster | null | Episode-based reinforcement learning (ERL) algorithms treat reinforcement learning (RL) as a black-box optimization problem where we learn to select a parameter vector of a controller, often represented as a movement primitive, for a given task descriptor called a context. ERL offers several distinct benefits in compar... | [
"Movement Primitives",
"Episode-Based",
"Black-Box",
"Deep Reinforcement Learning",
"Trust Regions"
] | A novel algorithm for deep episode-based reinforcement learning | 225 | 2210.09622 | title_snapshot |
DE8rdNuGj_7 | LEADER: Learning Attention over Driving Behaviors for Planning under Uncertainty | https://openreview.net/forum?id=DE8rdNuGj_7 | [
"Mohamad Hosein Danesh",
"Panpan Cai",
"David Hsu"
] | Oral | null | Uncertainty in human behaviors poses a significant challenge to autonomous driving in crowded urban environments. The partially observable Markov decision process (POMDP) offers a principled general framework for decision making under uncertainty and achieves real-time performance for complex tasks by leveraging Monte ... | [
"Planning under uncertainty",
"Integrating planning and learning",
"Autonomous driving"
] | We propose an algorithm that learns attention over human behaviors for planning under uncertainty. | 228 | 2209.11422 | title_snapshot |
RzhhFh4rkWu | Discriminator-Guided Model-Based Offline Imitation Learning | https://openreview.net/forum?id=RzhhFh4rkWu | [
"Wenjia Zhang",
"Haoran Xu",
"Haoyi Niu",
"Peng Cheng",
"Ming Li",
"Heming Zhang",
"Guyue Zhou",
"Xianyuan Zhan"
] | Poster | null | Offline imitation learning (IL) is a powerful method to solve decision-making problems from expert demonstrations without reward labels. Existing offline IL methods suffer from severe performance degeneration under limited expert data. Including a learned dynamics model can potentially improve the state-action space co... | [
"Offline Imitation Learning",
"Model-based Learning"
] | Model-based offline imitation learning for small datasets | 229 | 2207.00244 | title_snapshot |
nuAGobCwb8V | Representation Learning for Object Detection from Unlabeled Point Cloud Sequences | https://openreview.net/forum?id=nuAGobCwb8V | [
"Xiangru Huang",
"Yue Wang",
"Vitor Campagnolo Guizilini",
"Rares Andrei Ambrus",
"Adrien Gaidon",
"Justin Solomon"
] | Poster | null | Although unlabeled 3D data is easy to collect, state-of-the-art machine learning techniques for 3D object detection still rely on difficult-to-obtain manual annotations. To reduce dependence on the expensive and error-prone process of manual labeling, we propose a technique for representation learning from unlabeled Li... | [
"Representation learning",
"object detection",
"point cloud sequences"
] | Designing self-supervised tasks using unlabeled point cloud sequences to improve performance of object detection task | 232 | null | null |
4nt6RUGmILw | i-Sim2Real: Reinforcement Learning of Robotic Policies in Tight Human-Robot Interaction Loops | https://openreview.net/forum?id=4nt6RUGmILw | [
"Saminda Wishwajith Abeyruwan",
"Laura Graesser",
"David B D'Ambrosio",
"Avi Singh",
"Anish Shankar",
"Alex Bewley",
"Deepali Jain",
"Krzysztof Marcin Choromanski",
"Pannag R Sanketi"
] | Oral | null | Sim-to-real transfer is a powerful paradigm for robotic reinforcement learning. The ability to train policies in simulation enables safe exploration and large-scale data collection quickly at low cost. However, prior works in sim-to-real transfer of robotic policies typically do not involve any human-robot interaction ... | [
"sim-to-real",
"human-robot interaction",
"reinforcement learning"
] | We present a method for iteratively learning human behavior models for effective sim-to-real learning of robotic policies for human-robot interaction tasks. | 235 | 2207.06572 | title_snapshot |
t-IO7wCaNgH | Learning Temporally Extended Skills in Continuous Domains as Symbolic Actions for Planning | https://openreview.net/forum?id=t-IO7wCaNgH | [
"Jan Achterhold",
"Markus Krimmel",
"Joerg Stueckler"
] | Oral | null | Problems which require both long-horizon planning and continuous control capabilities pose significant challenges to existing reinforcement learning agents. In this paper we introduce a novel hierarchical reinforcement learning agent which links temporally extended skills for continuous control with a forward model in ... | [
"temporally extended skill learning",
"hierarchical reinforcement learning",
"diverse skill learning"
] | We jointly learn continuous control skills and a forward model of their effect in a symbolic abstraction, which can be leveraged for planning and plan execution | 236 | 2207.05018 | title_snapshot |
00OqOl4txhe | Learning Interpretable BEV Based VIO without Deep Neural Networks | https://openreview.net/forum?id=00OqOl4txhe | [
"Zexi Chen",
"Haozhe Du",
"Xuecheng XU",
"Rong Xiong",
"Yiyi Liao",
"Yue Wang"
] | Poster | null | Monocular visual-inertial odometry (VIO) is a critical problem in robotics and autonomous driving. Traditional methods solve this problem based on filtering or optimization. While being fully interpretable, they rely on manual interference and empirical parameter tuning. On the other hand, learning-based approaches all... | [
"VIO",
"Interpretable Learning"
] | A fully differentiable, and interpretable, BEV based VIO model for robots with local planar motion that can be trained without deep neural networks. | 238 | 2109.12292 | title_snapshot |
7CrXRhmzVVR | Solving Complex Manipulation Tasks with Model-Assisted Model-Free Reinforcement Learning | https://openreview.net/forum?id=7CrXRhmzVVR | [
"Jianshu Hu",
"Paul Weng"
] | Poster | null | In this paper, we propose a novel deep reinforcement learning approach for improving the sample efficiency of a model-free actor-critic method by using a learned model to encourage exploration. The basic idea consists in generating artificial transitions with noisy actions, which can be used to update the critic. To co... | [
"Reinforcement learning",
"Data augmentation",
"Imaginary exploration",
"Optimistic initialization"
] | Exploratory actions in learned model with optimistically-initialized critic improve value estimation and policy performance. | 240 | null | null |
uUxDTZK3o3X | Robustness Certification of Visual Perception Models via Camera Motion Smoothing | https://openreview.net/forum?id=uUxDTZK3o3X | [
"Hanjiang Hu",
"Zuxin Liu",
"Linyi Li",
"Jiacheng Zhu",
"Ding Zhao"
] | Poster | null | A vast literature shows that the learning-based visual perception model is sensitive to adversarial noises, but few works consider the robustness of robotic perception models under widely-existing camera motion perturbations. To this end, we study the robustness of the visual perception model under camera motion pertur... | [
"Certifiable Robustness",
"Camera Motion Perturbation",
"Robotic Perception"
] | We study the robustness of the visual perception model under camera motion perturbations and we propose a motion smoothing technique to give a certification guarantee under camera motion perturbations for any black-box image classification models. | 241 | 2210.04625 | title_snapshot |
vyR0C7zvozS | GLSO: Grammar-guided Latent Space Optimization for Sample-efficient Robot Design Automation | https://openreview.net/forum?id=vyR0C7zvozS | [
"Jiaheng Hu",
"Julian Whitman",
"Howie Choset"
] | Poster | null | Robots have been used in all sorts of automation, and yet the design of robots remains mainly a manual task. We seek to provide design tools to automate the design of robots themselves. An important challenge in robot design automation is the large and complex design search space which grows exponentially with the numb... | [
"Robot Design Automation",
"Latent Optimization",
"Graph Grammar"
] | A framework for sample-efficient robot design automation through unsupervised learning and Bayesian Optimization. | 243 | 2209.11748 | title_snapshot |
Bf6on28H0Jv | Masked World Models for Visual Control | https://openreview.net/forum?id=Bf6on28H0Jv | [
"Younggyo Seo",
"Danijar Hafner",
"Hao Liu",
"Fangchen Liu",
"Stephen James",
"Kimin Lee",
"Pieter Abbeel"
] | Poster | null | Visual model-based reinforcement learning (RL) has the potential to enable sample-efficient robot learning from visual observations. Yet the current approaches typically train a single model end-to-end for learning both visual representations and dynamics, making it difficult to accurately model the interaction between... | [
"Visual model-based RL",
"World models"
] | We present a visual model-based RL framework that decouples visual representation learning and dynamics learning. | 245 | 2206.14244 | title_snapshot |
K8W6ObPZQyh | On-Robot Learning With Equivariant Models | https://openreview.net/forum?id=K8W6ObPZQyh | [
"Dian Wang",
"Mingxi Jia",
"Xupeng Zhu",
"Robin Walters",
"Robert Platt"
] | Poster | null | Recently, equivariant neural network models have been shown to improve sample efficiency for tasks in computer vision and reinforcement learning. This paper explores this idea in the context of on-robot policy learning in which a policy must be learned entirely on a physical robotic system without reference to a model,... | [
"Manipulation",
"Reinforcement Learning",
"Equivariance"
] | This paper demonstrates sample-efficient on-robot learning in manipulation using equivariant models. | 246 | 2203.04923 | title_snapshot |
GTyBkq36tjx | Neural Geometric Fabrics: Efficiently Learning High-Dimensional Policies from Demonstration | https://openreview.net/forum?id=GTyBkq36tjx | [
"Mandy Xie",
"Ankur Handa",
"Stephen Tyree",
"Dieter Fox",
"Harish Ravichandar",
"Nathan D. Ratliff",
"Karl Van Wyk"
] | Poster | null | Learning dexterous manipulation policies for multi-fingered robots has been a long-standing challenge in robotics. Existing methods either limit themselves to highly constrained problems and smaller models to achieve extreme sample efficiency or sacrifice sample efficiency to gain capacity to solve more complex tasks w... | [
"Imitation Learning",
"Dexterous Manipulation"
] | We contribute a structured approach for sample-efficient learning of dexterous manipulation skills for a 23 DoF physical hand-arm system from demonstrations, by leveraging Geometric Fabrics, a recent theoretical framework for robot motion generation. | 247 | null | null |
sygvGP-YLfx | See, Hear, and Feel: Smart Sensory Fusion for Robotic Manipulation | https://openreview.net/forum?id=sygvGP-YLfx | [
"Hao Li",
"Yizhi Zhang",
"Junzhe Zhu",
"Shaoxiong Wang",
"Michelle A Lee",
"Huazhe Xu",
"Edward Adelson",
"Li Fei-Fei",
"Ruohan Gao",
"Jiajun Wu"
] | Poster | null | Humans use all of their senses to accomplish different tasks in everyday activities. In contrast, existing work on robotic manipulation mostly relies on one, or occasionally two modalities, such as vision and touch. In this work, we systematically study how visual, auditory, and tactile perception can jointly help robo... | [
"Multisensory Perception",
"Robotic Manipulation",
"Robot Learning"
] | We use vision, audio, and touch on two challenging robotic manipulation tasks---dense packing and pouring, and demonstrate the benefit of multisensory perception in robot learning. | 250 | 2212.03858 | title_snapshot |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.