Computer Vision Datasets

This dataset contains 87,871 face images with annotations for 106 facial landmarks, includes yellow, black, white and Indian races. 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 models, face recognition systems, and facial landmark detection algorithms in AI and computer vision.

This dataset includes 3,010 images of 602 people across multiple races. The semantic segmentation area includes headphones, glasses, body and background.This dataset can be used for training AI models in human body segmentation, video conference behavior detection, and smart vision applications.

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 diversity includes multiple scenes, light conditions, ages, shooting angles, and poses. In terms of annotation, we adpoted instance segmentation annotations on human body. 22 landmarks were also annotated for each human body. This dataset is suitable for training and benchmarking AI models in human body analysis, gesture recognition, and multi-person detection.

This dataset contains 116,048 sets of 3D handpose data, each set includes hand mask image(RGB, 24-bit), depth image(16-bit), camera intrinsic parameter file(TXT), 3D keypoints file(OBJ), mesh file(OBJ), gesture type file(TXT), keypoints demo image(JPG), and mesh demo image(JPG). The data is collected indoors, with the right hand (no handheld objects), covering both first-person and third-person perspectives, multiple gesture types, finger poses, hand overall rotation poses, individuals and Kinect devices used. This dataset does not include personally identifiable facial information, with hand mask images and depth images aligned. This dataset can be used for tasks such as handpose recognition, hand 3D reconstruction, and hand keypoints detection.

This dataset includes 9,000 images from 180 individuals, each labeled with 21 hand gesture landmarks along with attributes such as visibility, gesture type, gender, age, nationality, and vehicle type. Captured across diverse driving environments and time periods, this data supports research and model training in gesture recognition, pose estimation, and driver behavior analysis. It is especially useful for developing driver monitoring systems (DMS) and in-cabin interaction technologies. All data is compliant with GDPR, CCPA, and PIPL regulations and has been quality verified by leading AI companies.

This dataset includes multiple ages, multiple time periods and behaviors (dangerous behaviors, fatigue behaviors, visual movement behaviors). In terms of annotation, 72 facial landmarks (including pupils), face attributes, gesture bounding boxes, seatbelt bounding boxes, pupil landmarks and behavior categories were annotated in the data. This data can be used for tasks such as driver behavior analysis, fatigue detection, attention monitoring, and training AI models for intelligent in-cabin systems.

This dataset contains 142,636 images of 41 camera brands, collected from both indoor and outdoor scenes. The data diversity includes multiple scenes, multiple camera brands, different time periods and different image resolutions. This dataset can be used for tasks such as camera brand recognition, logo detection, image understanding, and scene classification.

This dataset contains 43,411 high-quality trademark images spanning 464 categories. The collecting environments include indoor and outdoor scenes. in this dataset, the image is clear without watermark, each image contains at least one trademark. The dataset can be used for trademark recognition, classification, brand logo detection, and other computer vision applications.