ShareGPT4Video: video captions
4.8M multimodal GPT-4-Vision video descriptions for LVLM and text-to-video models.
- Visual question answering
- Question answering
- English
Hugging Face datasets with plain-language descriptions: what's inside and what models to train on them. 4,788 in the catalog so far, with new ones added every hour.
Answered by Gemini 4 · the chat is kept only in this tab
4.8M multimodal GPT-4-Vision video descriptions for LVLM and text-to-video models.
Synchronized video and tactile glove readings to transfer human fine motor skills to robots.
An archive of 500 full broadcasts of European football matches and derived data for computer vision.
4M robotic clips with RGB, depth, and optical flow for image-to-video.
Updated benchmark with English videos and annotations for comprehensive evaluation of video models.
37.6 thousand text-video pairs from 11 generators with human annotations for training evaluation models.
A large collection of labeled images and reference videos in English to check consistency and physical plausibility of scene generators.
8.43 TB of synchronized sensor data for autonomous driving and robotics.
A large dataset of human action videos and images for motion modeling and robot pre-training.
48k screencasts (12.3k hours) for training and evaluating video and robotic models.
A massive video corpus with text descriptions and tabular metadata from the decentralized OMEGA Labs network.
Metadata, indexes, and camera-aligned video reconstruction packages for training world models.
Large open dataset of video pairs with instructions for training video editing models.
A benchmark for audio-video deepfake detection across multiple languages.
11,980 English video examples for gesture classification and sign language recognition.
100 million to 1 billion simulator-generated examples for training robotic foundation models without teleoperation.
Archive of saved YouTube videos for the well-known RealEstate10K camera trajectory dataset.
A set of 1,500 driving videos with accident and normal situations for training collision prediction models.
1,500 Nexar dashcam video clips with accidents and normal driving for collision avoidance systems.
Short text snippets from English video tutorials with timestamps for search and RAG.
Millions of records from a multimodal corpus of natural conversations between strangers via video call.
A robotics multimodal dataset featuring hand motion recordings and applied force in real-world tasks.
Synchronized audio, MIDI, motion capture, and fingering data for piano performance analysis.
Participant audio, video, and texts in English and Chinese with clinically confirmed diagnoses for distinguishing disorders.