MMMich commited on
Commit
9148e52
·
verified ·
1 Parent(s): 527f1f4

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +79 -0
README.md CHANGED
@@ -1,3 +1,82 @@
1
  ---
2
  license: mit
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: mit
3
+ tags:
4
+ - robotics
5
+ - in-context-learning
6
  ---
7
+ # Semantic-Discrimination-Test-Set
8
+
9
+ This repository releases the test resources of the paper [“In-context Robot Learning Made Simple: A Democratized Recipe for Manipulation Tasks”](https://arxiv.org/abs/2609.38173)
10
+
11
+ ## Semantic Discrimination Test Set
12
+
13
+ The test data is collected via our tailored ICL-oriented data pipeline, which is not the official RoboTwin setting. The archive `ICL-oriented-training-set.tar.gz` is extracted to obtain the training set organized into LeRobot `v2.1` directory structure.
14
+
15
+ ### Directory structure
16
+
17
+ Our ICL-oriented data pipeline produces episodes that are organized as:
18
+
19
+ ```text
20
+ <scene_name>/
21
+ placement0/ # One scene -> Multi-placement
22
+ <task0_name>/ # One placement -> Multi-task
23
+ <embodiment>[__pert0]/ # One task -> Multi-perturbation
24
+ data/
25
+ episode<I>.hdf5
26
+ instructions/
27
+ episode<I>.json
28
+ video/
29
+ episode<I>.mp4 # One perturbation -> Multi-visual-variant
30
+ scene_info.json # Task-level meta info
31
+ <embodiment>[__pert1]/
32
+ ...
33
+ <task1_name>/
34
+ ...
35
+ placement_meta.json # Placement-level meta info
36
+ placement1/
37
+ ...
38
+ completion.json # Collection completion markup
39
+ manifest.json # Scene configuration
40
+ ```
41
+
42
+ In this test set, each test episode is reorganized as:
43
+
44
+ ```text
45
+ <task_name>__<serial_number>/
46
+ placement0/
47
+ <task_name>/<embodiment>
48
+ data/
49
+ instructions/
50
+ video/
51
+ scene_info.json
52
+ placement_meta.json
53
+ completion.json
54
+ manifest.json
55
+ ```
56
+
57
+ This removes all tasks unrelated to the evaluation, saves storage space, and still allows the framework designed for the aforementioned data structures to read the data.
58
+
59
+ ## Project
60
+
61
+ - Project page: https://simpleicl.github.io/simpleicl/
62
+ - Paper: https://arxiv.org/abs/2609.38173
63
+ - Code repository: https://github.com/simpleicl/simple-icl
64
+
65
+ ## Citation
66
+
67
+ ```bibtex
68
+ @article{li2026context,
69
+ title={In-context Robot Learning Made Simple: A Democratized Recipe for Manipulation Tasks},
70
+ author={Li, Minxing and Han, Minghao and Zhao, Weizhi and Wang, Hanwen and Liu, Xiangshuo and Shang, Shuyao and Zhou, Jingxiang and Sun, Mingchao and Pan, Hongyu and Xu, Mu and others},
71
+ journal={arXiv preprint arXiv:2609.38173},
72
+ year={2026}
73
+ }
74
+ ```
75
+
76
+ ## License
77
+
78
+ This release is a derivative of RoboTwin 2.0 framework for ICL-oriented data collection.
79
+
80
+ It is released under the MIT license, consistent with the upstream RoboTwin 2.0 release. Please also refer to the original RoboTwin project and dataset for upstream attribution and terms.
81
+
82
+ > RoboTwin repository: https://github.com/RoboTwin-Platform/RoboTwin