Computer Vision Datasets

1,056 People Living_face & Anti-Spoofing Data. The collection scenes include indoor and outdoor scenes. The data includes male and female. The age distribution ranges from juvenile to the elderly, the young people and the middle aged are the majorities. The data includes multiple postures, multiple expressions, and multiple anti-spoofing samples. The data can be used for tasks such as face payment, remote ID authentication, and face unlocking of mobile phone.

23,110 People Multi-race and Multi-pose Face Images Data. This data includes Asian race, Caucasian race, black race, brown race and Indians. Each subject were collected 29 images under different scenes and light conditions. The 29 images include 28 photos (multi light conditions, multiple poses and multiple scenes) + 1 ID photo. This data can be used for face recognition related tasks.

87,871 Images of 106 Facial Landmarks Annotation Data (complicated scenes)，this dataset includes yellow race, black race, white race and Indian people. In order to be more challenging, the data includes multiple scenes, multiple poses, different ages, light conditions and complicated expressions. This data can be used for tasks such as face detection and face recognition.

1,196 People Multi-race and Multi-pose Face Images & Videos Data, 22 images and 40 videos per person. The collection environment includes indoor and outdoor scenes. This data can be used for face recognition and other tasks.

602 People –3,010 Images Multi-Races Human Body Semantic Segmentation Data,The data diversity includes headphones, body, background,and glasses.In terms of annotation, we adpoted segmentation annotations on headphones, body, background and glasses.The data can be used for tasks such as human body segmentation and the behavior detection of Video conference.

This dataset features 30 million high-resolution video clips sourced from authorized and legally compliant channels. Each video offers exceptional clarity, color accuracy, and scene diversity across various environments and themes. All clips include clearly documented copyright ownership and commercial usage rights, making them safe and reliable for both academic research and business applications. The dataset is ideal for computer vision tasks such as video classification, action recognition, multimodal learning, object tracking, and content generation. Whether you're training deep learning models or sourcing clean, scalable visual data, this video dataset offers a comprehensive, large-scale solution.

This dataset contains 80 million high-quality vector images (SVG, EPS, AI formats), offering a vast collection for use in computer vision, machine learning, and creative applications. Each image is copyright-cleared and legally sourced through authorized channels, with transparent usage rights for both commercial and academic purposes. The dataset features a wide variety of vector content—icons, illustrations, infographics, and more—with excellent color fidelity and scalable resolution. Ideal for AI model training (e.g., image classification, object recognition), generative design models, and creative design inspiration, this resource ensures traceable IP rights and enables safe, large-scale usage in real-world environments.

This 3D high-fidelity synthetic dataset simulates real-world Driver Monitoring System (DMS) environments using photorealistic 3D scene modeling. It includes multi-modal sensor outputs such as camera images, videos, and point clouds, all generated through simulation. The dataset is richly annotated with object classification, detection, and segmentation labels, as well as human pose data (head, eye, arm, and leg position/orientation), camera parameters, and temporal metadata such as illumination and weather conditions. Ideal for training and evaluating models in autonomous driving, robotics, driver monitoring, computer vision, and synthetic perception tasks.

This dataset contains 2,981 pairs of foggy and clear images collected from a wide range of real-world outdoor scenes, including urban streets, rural roads, mountains, parks, ruins, buildings, shops, flowers, trees, and seaside views. Each image pair captures the same scene under foggy and no-fog conditions, with variations in time of day, viewpoint, and fog density. This dataset is ideal for tasks such as single image dehazing, fog removal, image enhancement in adverse weather conditions, and computer vision robustness testing. All images are high-quality and suitable for training deep learning models in real-world defogging applications.