Loading...
Loading...
Image classification, object detection, segmentation, face recognition, OCR, image generation, video understanding
16,095 datasets
ArkAiLab-Adl developed and maintains the Nexora Vision Dataset v1 Large, a large-scale collection of images for AI research. The dataset contains a diverse collection of HD, 2K, and 4K images spanning multiple visual styles, including cinematic, realistic, anime, and sci-fi. It was last updated on February 19, 2026.
Global trends in governance, civil liberties, and political rights from 2006 to 2024. The dataset appears to track democracy and freedom metrics over time, with a title indicating coverage may extend back to 1972. It is hosted on Kaggle, but the author, organization, and specific data structure are unknown.
The dataset is a directory for the municipal enterprise 'Haisyn Housing and Maintenance Office' (KP 'Haisynska ZHEK') of the Haisyn City Council. It is provided via the eu_open_data platform and was last updated on March 3, 2026. The specific contents and scale of the directory are not detailed in the provided metadata.
mCellSeg is a dataset for microscopy cell instance segmentation. The dataset was sourced from Kaggle, but the author, organization, and specific collection details are unknown. Its size, row count, and last update date are unspecified.
A dataset named 'cnn_arxiv_mediasum-qwen3-1.7b-epoch100-lr0.1-val32' published on Kaggle. The title suggests it contains text data, likely for training or evaluating a language model. Its specific content and scale require verification after download.
A collection of image embeddings likely derived from handwriting samples using Convolutional Neural Networks (CNN) and Gradient Energy Image (GEI) methods. The dataset title suggests it is based on version 24.4 and may contain 225-dimensional feature vectors. It is published on Kaggle, but the author, organization, and specific data collection details are unknown.
A text dataset likely containing summarized news articles, derived from sources such as CNN, ArXiv, and MediaSum. It appears to be processed or generated by a machine learning model named Qwen3-1.7B, trained for 50 epochs with a learning rate of 0.1. The dataset is hosted on Kaggle.
Yossef Bodansky's book 'The Secret History of the Iraq War' analyzes the 2003 Iraq War and its aftermath, drawing on previously untapped intelligence and regional sources. The work presents an account of intelligence failures, Saddam Hussein's alleged ties to terrorist groups, and the origins of the anti-U.S. insurgency. It was sourced from the author's role as director of the Congressional Task Force on Terrorism and Unconventional Warfare.
East German archival sources, including previously secret documents from the Stasi and Communist Party, provide a behind-the-scenes view of Cold War Germany during the era of détente. The work by M.E. Sarotte explores the motives of the GDR and its Soviet backers in responding to West German Ostpolitik and U.S. foreign policy under Nixon. It focuses on public and secret contacts between East and West Germany during Brandt's chancellorship, exposing negotiation tactics and superpower maneuvering influenced by concerns over China.
The book 'Good Americans' examines the contributions of thousands of Italian and Jewish immigrants to the U.S. war effort during World War I. It analyzes their roles in relief, recruitment, conservation, and war bond campaigns, as well as their service as soldiers on the Western Front. The work argues these experiences accelerated the political transformation of first and second-generation immigrants in the United States.
Recently declassified intelligence documents reveal America's knowledge of and collaboration with the German resistance to Hitler during World War II. The collection includes reports from the OSS, the CIA's predecessor, detailing internal German society and specific plots like the July 20th assassination attempt. Documents are arranged in chronological order and accompanied by explanatory footnotes and a comprehensive introduction.
1914-1919 data on atrocity propaganda from the paperswithcode platform. The dataset appears to be closed-license and its specific contents, such as column definitions and sample data, are unavailable. The description provided discusses cultural history topics, which may indicate a mismatch or error in the metadata.
CMS Data provides financial information for hospitals in Michigan. The dataset likely contains metrics related to hospital operations and finances. The temporal coverage spans from 2018 to 2023.
Global real-time Earth system data sourced from all World Meteorological Organisation Member countries and territories. The Met Office and NOAA operate this Global Cache service, curating and publishing the dataset on behalf of the WMO. Data is retained for a 24-hour period.
A continuous, real-time monitoring project of global broadcast, print, and web news from nearly every country in over 100 languages. It identifies people, locations, organizations, themes, sources, emotions, quotes, images, and events driving global society. The dataset is hosted on AWS Open Data and its license terms are provided by the GDELT Project.
A labeled dataset of toy cars intended for training YOLOv8 OBB object detection models. The dataset likely contains images of toy cars annotated for oriented bounding box detection. The author, organization, and specific size are unknown.
A dataset titled 'keypoints_data' published on Kaggle. The dataset's specific content, size, and origin are not detailed in the provided metadata. Its title suggests it likely contains coordinate or feature point information, potentially for tasks like object detection or pose estimation.
Synthetic records totaling between 100,000 and 1,000,000 simulate the monthly operational characteristics of Indonesian MSMEs (UMKM). Created by ZakyF and updated in March 2026, the data models business scale, operational efficiency, and customer sentiment. It provides a consistent framework for exploring the relationships between financial metrics and customer feedback.
YOLO-formatted images for detecting obstacles across varied environments. The dataset likely contains annotated street scenes and indoor settings. Its author, organization, and size are unspecified.
Organ-Agents Repro Code Full-v1 is a dataset published on Kaggle. The title suggests it relates to computer vision and reproducibility for an 'Organ-Agents' project. The dataset likely contains code and associated materials intended for replicating research or experiments.