HOT3D clips collection for 3D object and hand pose estimation
Curated video fragments and fisheye camera images from the HOT3D benchmark for 3D vision tasks.
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
Curated video fragments and fisheye camera images from the HOT3D benchmark for 3D vision tasks.
17.3M images, 24.3M examples, and 88.9M turns for training modern open vision-language models.
A dataset of road images in clear weather, fog, rain, and snow for training perception systems.
Classic ILSVRC 2012 image dataset organized by the WordNet hierarchy for training and evaluating classifiers.
Benchmark for multimodal understanding and reasoning featuring questions with images across multiple disciplines.
Road scene image dataset for perception research.
114k 3D head and neck MRI scans from 800 open datasets for pretraining medical models.
Multimodal dataset with images and English/Chinese questions for Chinese porcelain appraisal.
A dataset of RGB maritime vessel images translated into a synthetic infrared range using the FLUX model.
An English set of 1k–10k examples for evaluating models on video question-answering tasks.
Large curated dataset of 103.8 million image-text pairs for text-to-image training.
1 to 10 thousand real bridge images with multi-label segmentation across 19 damage and structural classes.
3D CT abdominal benchmark for Visual Question Answering with 438K QA pairs and clinical subtypes.
7 synthetic 3D scenes with complex lighting effects and dozens of viewpoints for relighting tasks.
7,953 ABO product 3D models in GLB with PBR textures, real-world dimensions, and browser-optimized size.
English visual benchmark to test MLLM ability to understand 3D scenes from two different viewpoints.
Synchronized ground and aerial road images with camera parameters for 3D reconstruction and neural rendering.
100K to 1M synthetic infrared ship images with artificial degradation.
Multimodal benchmark of 2500 difficult questions across dozens of academic topics.
Corpus of texts and images from gov.br government news.
Benchmark from ServiceNow with screenshots and long computer agent action chains to evaluate judge models.
Large dataset of images and ratings for training models to predict aesthetic rendering quality.
Over 3.5M training examples with table images, HTML markup, and data for 17 different tasks.
Kaggle AnimalCLEF26 competition dataset mirror with animal images and metadata for classification.