Loading...
Loading...
Image classification, object detection, segmentation, face recognition, OCR, image generation, video understanding
16,281 datasets
A dataset likely containing images and annotations for training object detection models using the YOLO (You Only Look Once) framework. It appears to be related to the YOLO-11n and YOLO-6-scaled architectures, suggesting a focus on model variants and scaling. The dataset is hosted on Kaggle, but its specific content, size, and creation details are not provided.
The dataset tracks sidewalk defect violations and related inspections in New York City, managed by the City of New York's Department of Transportation. It includes records for violations, inspections, and repinspections, with data last updated in March 2026.
Labeled images depict faults and normal conditions in machinery components. The dataset is hosted on Kaggle, but details about its author, organization, and creation date are unknown. The total number of images, file formats, and specific component types are not specified in the provided metadata.
Food price data from Central Java, Indonesia, published on Kaggle. The dataset likely contains price information for various food commodities. Specific details such as the number of records, time range, and data collection methodology are not provided in the metadata.
ShrimpDiseaseImageBD is an image dataset published on Kaggle. The title suggests it contains images for training YOLO models to detect shrimp diseases. Metadata is minimal; actual content requires verification after download.
Insurance claim images from the United States processed with optical character recognition. The description indicates the dataset includes detailed metadata for each image. The original source, size, and specific content of the images are not detailed.
Organic-CoT provides 195,000 high-fidelity synthetic examples designed to teach models expert-like reasoning processes. The dataset was created by CodonProject and published on HuggingFace in February 2026. It focuses on trial-and-error, reflection, and strategic planning rather than rigid step-by-step outputs.
British Geological Survey data documents synthetic groundwater microcosm experiments measuring biodegradation of the biocide glutaraldehyde. The dataset captures the effects of additives like salt, guar gum, and polyacrylamide on microbial degradation processes from the UKGEOS Glasgow Observatory.
MedVision is the official release of a dataset for medical vision-language model research. The dataset was created by YongchengYAO and was last updated on March 2, 2026. Its specific size and content details are not provided in the available metadata.
Scanned PDF documents from the Ministry of Finance (MOF) of Nepal, intended for optical character recognition (OCR) processing. The dataset is hosted on Kaggle, but its specific size, date of collection, and the number of documents are unknown. The content likely contains official government forms, reports, or financial documents.
A synthetic dataset likely designed for training YOLO object detection models. The dataset appears to be balanced, which suggests an effort to mitigate class imbalance common in computer vision tasks. It is hosted on Kaggle, but the author, organization, and specific creation details are unknown.
A dataset titled 'Tomato Leaf Diseases.v3iyolov8' is hosted on Kaggle. The dataset likely contains images of tomato leaves for computer vision tasks, as suggested by its title and the 'v3iyolov8' suffix, which may indicate a version or model association. No further metadata on size, source, or annotation details is provided.
Modele CNN is a dataset published on Kaggle. The title suggests it contains data related to a Convolutional Neural Network model. Its specific content, scale, and origin are unknown from the provided metadata.
COCO VGG Features likely contains pre-computed visual feature vectors from the Common Objects in Context (COCO) dataset. The title suggests the features were extracted using a VGG convolutional neural network architecture, a common practice for providing ready-to-use inputs for machine learning models. This dataset is hosted on Kaggle, but detailed metadata about its size, columns, and creation date is unavailable.
Wikipedia Image Captioning Romanian is a dataset hosted on Kaggle. The title suggests it likely contains images paired with descriptive text captions in the Romanian language. The specific scale, source, and creation details are not provided in the available metadata.
A dataset titled 'coco-vgg-features-chunk1' published on Kaggle. The name suggests it contains precomputed visual features, likely extracted using a VGG convolutional neural network from images in the COCO (Common Objects in Context) dataset. The dataset's author, organization, size, and specific contents are unknown from the provided metadata.
A dataset of precomputed visual features, likely derived from a VGG neural network model applied to images from the COCO (Common Objects in Context) dataset. The title suggests it is the third chunk of a larger collection. It is hosted on Kaggle, but the author, size, and specific feature dimensions are unknown.
YOLO PCB MULTIVIEW DATASET is a computer vision dataset hosted on Kaggle. The dataset likely contains images of printed circuit boards (PCBs) from multiple viewpoints, intended for training object detection models. The specific number of images, annotation details, and creation date are unknown from the provided metadata.
Kaggle hosts this dataset titled 'coco-vgg-features-valtest'. The dataset likely contains precomputed feature vectors extracted from images in the COCO (Common Objects in Context) validation and test sets using a VGG convolutional neural network. The specific number of rows, file formats, and author details are unknown from the provided metadata.
Medariana, Michigan and Nevada water temperature and other environmental data collected by NOAA NCEI. The dataset covers the period from September 22, 1976, to August 14, 1977. It is published on the nasa_earthdata platform.