Datasets
634 results
Video
GoPro-Raw-Videos
Raw GoPro Videos for Four Robotic Manipulation Tasks Project Page Paper Code Models Processed Dataset This repository contains…
10K–100K·MIT
Video
Kenyan Animal Behavior Recognition (KABR) Mini-Scene Raw Videos
Kenyan Animal Behavior Recognition (KABR) Mini-Scene Raw Videos
1M–10M·CC0
MultimodalSensor / Time-seriesTabular
2025-challenge-demos
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase version": "v2.1", "robot type": "R1Pro", "total episodes":…
100M–1B·MIT·Parquet
MultimodalSensor / Time-seriesTabular
aloha_sim_transfer_cube_human
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase version": "v2.0", "robot type": "aloha", "total episodes":…
10K–100K·MIT·Parquet
MultimodalTextVideo
phyground
PhyGround Project Page GitHub Paper PhyGround is a criteria-grounded benchmark for evaluating physical reasoning in video generation. The…
<1K·JSON
Video
PhysicalAI-SmartSpaces
Physical AI Smart Spaces Dataset Overview Comprehensive, annotated dataset for multi-camera tracking and 2D/3D object detection. This dataset…
1K–10K·CC-BY
MultimodalTextVideo
RoboCerebra_Progress_Estimation
RoboCerebra VideoQA Benchmark This directory contains the Video Question Answering benchmark for evaluating RoboCerebra models on action progress…
<1K·MIT·JSON
MultimodalTextVideo
ReactiveGWM-Datasets (SF2 / SF3)
ReactiveGWM-Datasets (SF2 / SF3)
10K–100K·CC-BY-NC·CSV
Video
FantasyPortrait-Multi-Expr
FantasyPortrait: Enhancing Multi-Character Portrait Animation with Expression-Augmented Diffusion Transformers 🔥 Latest News!! August 10, 2025: We…
10K–100K·Apache-2.0
Video
RoboMemArena
RoboMemArena 数据集文件元信息以及数据文件，请浏览“数据集文件”页面获取。 当前数据集卡片使用的是默认模版，数据集的贡献者未提供更加详细的数据集介绍，但是您可以通过如下GIT Clone命令，或者ModelScope SDK来下载数据集 下载方法 Task Categories…
<1K·Apache-2.0
Video
AgiBot-World-Beta-no-torso-movement
Dataset taken from BAAI-DataCube/agibot-lerobot-v30-217-task-collection, in LeRobot Datasets v3.0 format. Fisheye video features are dropped so there…
10K–100K·CC-BY-NC-SA
Video
PhysicTran38K
PhysicTran38K We provide 38K video-based dataset for physics-aware image editing, by casting editing as physical state transitions. To…
10K–100K·Apache-2.0