Weihan Xu PRO

WitneyWW

https://wx83.github.io/

wx83

AI & ML interests

None yet

Recent Activity

published a Space about 3 hours ago

WitneyWW/epic-kitchens-tactile-samples

updated a Space about 3 hours ago

WitneyWW/epic-kitchens-tactile-samples

published a Space about 5 hours ago

WitneyWW/holi-spatial-hdepic-walkthrough

View all activity

Organizations

None yet

WitneyWW 's Spaces 115

EPIC-KITCHENS tactile samples

✋

Explore cooking video clips and learn which tasks need touch

Holi-Spatial x HD-EPIC walkthrough

🧊

Convert egocentric video into 3D object boxes and spatial Q&A

Object Memory Eval

🎯

Human evaluation of object-centric memory annotations

HD-EPIC Question Pyramid — Gemini vs Qwen

🔺

All Gemini and Qwen runs on the HD-EPIC pyramid

ACTS Short-Horizon Predictions

🤖

Explore action‑conditioned short‑horizon video predictions

InstructAV2AV FT vs BASE (test200)

🎬

Edit videos to add or remove objects and sounds

ACTS Causal Rollouts

🤖

Explore predicted future video and sensor rollouts

JavisDiT fps-fixed — addition, clean-input cases

🔊

Add missing objects and sounds to your videos

JavisDiT fps-fixed results

📊

Generate videos with objects added, removed, and synced audio

ACTS visuo-tactile world model

🧪

Predict future visual and tactile video frames for robot tasks

Explore model evaluation metrics for robot vision data

Wrist ablation (mixed)

🤖

Compare model predictions with and without wrist stream

Wan5b wrist MoT val rollouts

🤖

Motherboard wrist MoT rollouts

🤖

Explore model rollout predictions for wrist camera data

Wrist MoT rollouts (React)

🤖

Explore model predictions with interactive videos and charts

AV-Highlight results

🔦

Highlight objects in video and boost their sound

JavisDiT 0904 checkpoint sweep

🎚

Explore video object add/remove results across checkpoints

JavisDiT 0904 ep027 large objects

🎬

Edit videos to add or remove objects via text prompts

JavisDiT 0904 ep027 vs noref ep017

🎬

Edit videos to add or remove objects by text prompt

avhl ropealign step6000 vs 6500

🎧

Compare video generation results from two models side‑by‑side

avhl RoPE-align sel20

🔦

Explore video highlights from different model checkpoints

MoT wrist-camera world model (val0909)

🤖

Explore model predictions and performance visualizations