Loading...
Loading...
Image classification, object detection, segmentation, face recognition, OCR, image generation, video understanding
17,006 datasets
Replication data for the paper 'Outsourcing of research and development and efficiency: a DEA non-parametric analysis of the contract research organisations industry' by Díaz and Sanchez-Robles (2023). The dataset was harvested from e-cienciaDatos and last updated on October 14, 2025.
Data from the OPTIMSOIL project measuring the Fraction of Photosynthetically Active Radiation intercepted by the canopy (fIPAR), incident PAR, transmitted PAR, intercepted PAR per subperiod, and Radiation Use Efficiency (RUE). The dataset was authored by Daniel Palmero Llamas and last updated on October 14, -2025.
A dataset of high-quality images with precise annotations for object detection tasks. It contains images for 33 different automotive brands, focusing on brand recognition in contexts like vehicles and logos. The dataset was created by haydarkadioglu and was last updated on 2025-09-24.
Imagenet21K Recaption is a dataset of approximately 13 million images across about 19 thousand classes, with labels provided as strings. The dataset was created by author gmongaras and was last updated on the Hugging Face platform on 2025-09-17. Images are stored in PNG format and can be decoded using the Python PIL library as shown in the description.
UniqueData's Grocery Store Receipts Dataset is a collection of photos from various grocery store receipts, designed for Optical Character Recognition tasks. Each image includes bounding box annotations for text segments categorized into four classes: item, store, date_time, and total. The dataset was last updated on October 1, 2025.
A curated subset of 25,000 images from a larger collection of over 1 million soccer images. The dataset was compiled by Adit-jain from multiple open-source sources, including the SoccerNet dataset, video game renders, and match footage, and was last updated on August 28, 2025.
Animepics is a pure image dataset in WebDataset format designed for pre-training anime-style models. The dataset is designed to be continuously updated, leveraging the features of WebDataset. It was authored by zenless-fab and last updated on October 2, 2025.
Serving as used for training sessions in a course on data refinement, cleaning, and anonymization. The course aims to teach principles of data refinement and the use of tools like OpenRefine and Amnesia. The dataset's specific row count, column count, and structure are not provided.
A large-scale, synthetically generated dataset for training and evaluating document understanding and Optical Character Recognition models. The dataset was created by FatimahEmadEldin and last updated on HuggingFace in September 2025. It provides page images paired with their corresponding structured HTML ground truth, with a focus on the Arabic language.
HESCAPE is a large-scale benchmark for multimodal learning in spatial transcriptomics, hosted by Peng-AI. The dataset is organized into multiple panel-specific configurations, including human-5k-panel, human-breast-panel, and human-colon-panel. It was last updated on September 16, 2025.
Cityscapes is a processed version of the original dataset containing 3,475 urban street scene images, converted to HuggingFace datasets format by tanganke. The dataset includes 2,975 training samples and 500 validation samples, with corresponding semantic segmentation labels and depth maps. It was last updated on the platform on September 7, -2025.
coco_synth_qwen3_next_0_1 is a synthetic image dataset containing 10010 images. The dataset was created by author testdummyvt using an automated pipeline to mirror the COCO80 object list, pairing each generated image with a short scene description and structured object entries. It was last updated on Hugging Face on 2025-09-14.
Exempt organization information is extracted monthly from the Internal Revenue Service’s Business Master File. The dataset provides records for Connecticut-based organizations, containing fields such as EIN, NAME, STATE, ASSET_AMT, INCOME_AMT, REVENUE_AMT, and NTEE_CD. It is updated nightly and maintained by the State of Connecticut.
This large-scale collection of multi-view video sequences focuses on the spatio-temporal localization of human-object interactions in retail settings. It provides synchronized camera perspectives to support the development of models for complex activity recognition and object manipulation tracking in real-world environments.
A combined dataset of approximately 25 million images from the ImageNet21K and CC12M collections, recaptioned by the author. The ImageNet21K portion contains about 13 million examples across roughly 19,000 classes, while the CC12M portion contains 12 million images created in 2021. The dataset was uploaded to Hugging Face by user 'gmongaras' in September 2025.
Pokrovsky City Council's executive committee education management organizational structure is documented. The dataset likely contains information about departments, roles, and staffing as defined by an organizational and administrative document. It was last updated on September 22, 2025, and originates from a Ukrainian government data portal.
Afri-Aya is a community-curated multilingual image dataset covering 13 major African languages with AI-powered categorization. It was created by the Cohere Labs Regional Africa community during the six-week Expedition Aya open-build challenge to include low-resource languages in vision-language tasks.
5,000 question-answer pairs demonstrating the Socratic method of teaching through guided questioning. The dataset was created by sanjaypantdsd and last updated on Hugging Face in September 2025. It has been cleaned to remove romantic and potentially inappropriate content.
Grefcoco is a benchmark dataset for Generalized Referring Expression (GREx) tasks, including segmentation, comprehension, and generation, developed by researcher henghuiding and presented at CVPR 2023. It provides a standardized framework for evaluating models on their ability to process natural language queries that may refer to multiple objects or no objects at all within an image.
Tri3D is a unified interface for accessing and processing major 3D autonomous driving datasets, including KITTI, nuScenes, and the Waymo Open Dataset. Developed by CEA-LIST and last updated in November 2025, it provides a standardized framework for handling multi-modal sensor data and 3D geometry.