Loading...
Loading...
Image classification, object detection, segmentation, face recognition, OCR, image generation, video understanding
16,273 datasets
A dataset published on Kaggle, likely containing images for training or evaluating YOLO (You Only Look Once) object detection models. The author, organization, and specific details such as size and license are unknown. The last update date is also unknown.
Olmocr Bench English is a document text extraction benchmark containing 6,483 English-only test cases across seven specialized categories including mathematical notation and complex layouts. Created by sarvamai as a filtered version of the allenai/olmOCR-bench dataset, it provides a targeted evaluation set for OCR models updated as of February 2026. The collection focuses on challenging document elements such as multi-column text, tables, and historical scans.
YOLO-CheckPoint is a dataset of model weights or checkpoints for the YOLO (You Only Look Once) object detection architecture, published on Kaggle. The dataset likely contains serialized model parameters for training or inference. Specific details on the number of checkpoints, their versions, and the training data used are unavailable.
A computer vision model for detecting camouflaged bears, likely built using the YOLOv8 architecture enhanced with the Convolutional Block Attention Module (CBAM). The dataset was published on Kaggle. The specific size, content, and creation details are not provided.
A dataset likely containing images annotated for object detection, as suggested by the title referencing YOLOv8, a popular computer vision model. It was published on Kaggle, but the specific number of images, annotation details, and creation date are unknown. The dataset appears to be a baseline model or benchmark for a task named 'Taco-4-5'.
SegOCR-Ensemble-A is a dataset published on Kaggle. The title suggests it contains data for training or evaluating an ensemble model for scene text detection and recognition (OCR). The dataset's specific content, size, and origin are not detailed in the available metadata.
SegOCR-Ensemble-B is a dataset published on Kaggle. Its title suggests it contains outputs from an ensemble model for scene text recognition, likely involving images with text. The dataset's specific content, size, and origin require verification after download.
best_yolo8_BDD100kv3 is a dataset hosted on Kaggle. The title suggests it is based on the BDD100K dataset, a large-scale driving scene dataset. Its specific content and scale require verification after download.
No_PGD_NoGAN_Noisy_Segment_ECGID_V3 is a dataset of electrocardiogram (ECG) signals, specifically version 3 of the ECGID database. The title suggests it contains noisy signal segments and was created without using PGD (Projected Gradient Descent) or GAN (Generative Adversarial Network) augmentation techniques. It is hosted on Kaggle, but detailed metadata about its size, structure, and origin is unavailable.
YOLO_5K is a dataset published on Kaggle. Its title suggests a focus on object detection, likely containing images annotated for use with YOLO models. The dataset's specific content, size, and origin are not detailed in the provided metadata.
ECGID_NOPGD_NOGAN_PER_RANDOM_SEGMENT_ATTENTION is a dataset published on Kaggle. Its title suggests it contains electrocardiogram (ECG) signals, likely processed with attention mechanisms for machine learning tasks. The dataset's specific content, scale, and origin are not detailed in the provided metadata.
A project by the British Geological Survey aims to develop a generic modeling engine for porosity development and flow in heterogeneous porous media, such as fractured chalk aquifers. The work focuses on generating porosity templates and investigating percolation and scaling phenomena for application in natural and anthropogenic scenarios.
A research project proposes to recalibrate the Palaeocene-Eocene boundary using Ar-Ar and U-Pb dating methods. The study, led by the British Geological Survey, aims to resolve discrepancies between magmatic event ages and accepted magnetostratigraphy time ranges. It involves dating sanidine from ash layers in northern Jutland and zircon/sphene from Danish and Hebridean ashes.
British Geological Survey data contains rhenium-osmium abundance and isotope measurements from Cretaceous stratigraphic sections in Hokkaido, Japan, and Sauzeries, France. The dataset supports analysis of the Ocean Anoxic Event 1a (OAE1a) driving mechanisms, as referenced in a 2024 Science Advances publication.
Parasites 1S07H is an object detection dataset containing 2,110 annotated microscopy images across 8 parasite species. The dataset is part of the Roboflow 100 benchmark, curated by LibreYOLO, and was last updated on Hugging Face in January 2026. It is split into 1,484 training, 411 validation, and 215 test images.
Eagle Ford Shale in South Texas provides a dataset of rhenium isotope compositions, concentrations, and associated geochemical measurements from bulk rock digestions. The data was collected to compare isotopic composition before and after oxidative weathering, using samples from drill core Innes-1 and outcrop sections DR5 and DR12. Measurements were obtained via MC-ICP-MS and Rock-Eval pyrolysis.
Replication data for a 2025 European Journal of International Relations study analyzes the relationship between rebel group grievances and the durability of peace agreements. Created by researchers Keels, Joo, and Wiegand, the dataset provides the empirical basis for survival analysis in conflict resolution.
Swansea University experimental data documents the synthesis, crystallinity, and CO2 capture performance of Ce-based metal-organic frameworks. The project was conducted from October to December 2021 at the Energy Safety Research Institute. Data includes powder X-ray diffraction (PXRD) results and CO2 sorption analysis for performance comparison.
Chemical composition data from bioleaching experiments on three Turkish karst bauxites. The experiments used organic acid, reductive, and oxidative bioleaching methods with uninoculated controls, conducted by the British Geological Survey to assess rare earth element recovery potential.
GEBench is a benchmark dataset from stepfun-ai for evaluating dynamic interaction and temporal coherence in Graphical User Interface generation. It addresses a gap in existing benchmarks by focusing on state transitions and temporal coherence within GUI-specific contexts. The dataset was last updated on February 25, 2026.