CADBench: Multimodal benchmark for CAD program generation
A multimodal benchmark of 18,000 samples to assist in CAD program generation and understanding.
- Image → text
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
A multimodal benchmark of 18,000 samples to assist in CAD program generation and understanding.
Panda robot manipulation episodes: 1,693 episodes, 273k frames, and 40 tasks for training robotics policies.
4.8M multimodal GPT-4-Vision video descriptions for LVLM and text-to-video models.
A complete set of COCO 2017 images and annotations for training and evaluating computer vision models.
Multimodal set of radiology scans with captions and medical concepts from PMC.
Problems from countries and competitions with images and text for QA and solution generation.
3,169 3D models with Python code in sand3d, OpenUSD export, and renders for text-to-3D and image-to-3D.
A large collection of labeled images and reference videos in English to check consistency and physical plausibility of scene generators.
Over 33 hours of road video, GPS, and IMU readings from a California highway for autopilot tasks.
A collection of images and texts for testing multimodal LLM robustness against jailbreak attacks.
Image and text connections from Wikipedia across many languages for training and evaluating multimodal retrieval.
8.43 TB of synchronized sensor data for autonomous driving and robotics.
HTML pages with screenshots and instructions for training web agents to perform tasks on websites.
High-resolution tissue images and gene expression data for biomedical models.
1K to 10K test examples with images and questions to evaluate vision capabilities in multimodal models.
A large dataset of human action videos and images for motion modeling and robot pre-training.
3D models of Tokyo districts and major Japanese cities in GLB format for streaming.
Over 7,500 synthetic road scene images with depth, segmentation, and object annotations for autonomous driving.
4,460 challenging medical questions spanning text and multimodal tasks with clinical images.
~3.3M web images with alt-text captions for image-to-text training.
Largest open human preference dataset for text-to-image models with 7.7M votes.
A benchmark of 789 images for training and evaluating dark photo brightening algorithms.
Compact image classification dataset with reduced resolution for fast experiments.
Dataset of source-to-stylized image pairs with text prompts for training style transfer.