Loading...
Loading...
Image classification, object detection, segmentation, face recognition, OCR, image generation, video understanding
15,926 datasets
RoboChallenge Table30 v2 Dataset includes 30 diverse manipulation tasks across 4 embodiments. The tasks, such as 'put_the_books_back' and 'tie_a_knot', involve object interaction and scene understanding. The dataset was created by RoboChallenge and was last updated on 2026-04-10.
SODA-D-YOLO is a dataset published on Kaggle. The title suggests it is intended for training or evaluating YOLO (You Only Look Once) object detection models. The dataset's specific contents, scale, and origin are not detailed in the available metadata.
CNN models likely trained from scratch for computer vision tasks. The dataset is hosted on Kaggle, a popular platform for data science competitions and projects. Specific details about the model architectures, training data, and performance metrics are not provided in the available metadata.
Keypoint_12_frames is a dataset published on Kaggle. The title suggests it contains image frames, likely from video, annotated with keypoints. The dataset's specific content, scale, and origin are not detailed in the available metadata.
MNIST is a classic dataset of handwritten digit images, likely intended for training and evaluating Convolutional Neural Networks (CNNs). It is hosted on the Kaggle platform. The specific version, size, and authorship details are not provided in the available metadata.
A widely recognized collection of handwritten digit images, likely intended for training and evaluating Convolutional Neural Networks (CNNs). The dataset is hosted on Kaggle, but specific details on its size, origin, and update history are not provided in the available metadata. Its content and structure must be verified after download.
A dataset named Keypoint_12_frames_multi_horizon, published on Kaggle. The title suggests it likely contains sequences of 12 frames for multi-horizon keypoint detection tasks. Metadata is minimal; actual content requires verification after download.
MNIST is a classic dataset of handwritten digit images, likely intended for training and evaluating Convolutional Neural Networks (CNNs). It is hosted on the Kaggle platform. The specific version, size, and authorship details are not provided in the available metadata.
Test Bengali OCR Dataset Small is a dataset published on Kaggle. Its title suggests it contains images of Bengali text and corresponding transcriptions for optical character recognition tasks. The dataset's specific size, collection method, and author are unknown from the provided metadata.
A dataset of images annotated for object detection, likely containing bees and wasps. It is hosted on Kaggle, but the specific collection date, author, and total number of images are unknown. The data appears to be formatted for training YOLO (You Only Look Once) models.
Tugas penambangan data translates to 'Data Mining Assignment'. The dataset is hosted on Kaggle, a popular platform for data science projects. Its specific content, size, and origin are not detailed in the provided metadata.
EgoCoT-Bench provides 3,172 samples across 351 unique videos for evaluating reasoning in first-person video. The benchmark, authored by DStardust and released in April 2026, is structured for evaluation with 300 public development samples and 2,872 public test samples. Its focus is on grounded and verifiable reasoning tasks within egocentric video contexts.
CERF Topline Figures provides high-level humanitarian indicators and summary statistics managed by the Central Emergency Response Fund. This CSV dataset, updated through March 2026, serves as the authoritative source for the organization's primary performance metrics on the Humanitarian Data Exchange.
Monthly uploads from the City of Austin's Community Registry list organizations receiving land development notices. The dataset enables contact of multiple registered groups by filtering on fields like Association Type or Association ZipCode. Organizations are registered to receive alerts for permit applications within 500 feet of their boundaries.
July to September 2019 sampling of surface water from 97 globally distributed river corridor systems. The dataset includes high-resolution dissolved organic matter characterization via FTICR-MS, NPOC, stable isotopes, anions, bacterial abundance, and DIC, produced by the Pacific Northwest National Laboratory. Data packages were updated multiple times through August 2023.
Over 1100 soil profiles were assembled to analyze changes in organic matter storage after land conversion. The data, authored by W. M. Post of Oak Ridge National Laboratory, shows an average carbon loss of 23% for soils with high initial carbon at 1-meter depth, while nitrogen loss averaged 6%. Regression analysis indicates carbon loss increases with initial soil storage and is influenced by the C:N ratio.
Per capita ethanol consumption data for persons aged 14+ in all U.S. states and Washington D.C. from 1977 to 2016. The dataset includes total consumption and separate figures for beer, wine, and spirits, originally compiled by the National Institute on Alcohol Abuse and Alcoholism. It was scraped from a PDF report and formatted for analysis by Jacob Kaplan.
Dataset OSD contains bottle and Conductivity-Temperature-Depth (CTD) data collected from multiple French and international platforms worldwide. Measurements support the World Ocean Circulation Experiment (WOCE) and Joint Global Ocean Flux Study (JGOFS) programs. Bottle parameters include dissolved oxygen, nitrate, nitrite, silicate, chlorophyll-a, dissolved organic carbon, and total phaeopigments, while CTD profiles capture temperature and salinity.
Temperature loggers deployed at Gannet Cay Reef collected sea water temperature data from 13 August 2014 to 07 November 2025. The data was aggregated by the Australian Ocean Data Network and last updated on the platform in March 2026. The specific number of loggers, sampling frequency, and data volume are not detailed in the provided metadata.
ImageNet-1K Animal Classes Mini 100 is a subset of the ImageNet-1K dataset focused on animal categories. It contains 100 images per class. The dataset was sourced from Kaggle, but the original author and license are unknown.