Type: Video Quantity: 50 hours Size: 190GB File Format: MP4, MOV Clarity: 1080x1920 (mostly), 720x1280 (some), 480x640 (few) Languages: Speech and music in Bulgarian, Russian and English Years of home videos capturing kids growing up, playing, and celebrating life's little moments. Includes parties, birthdays, events, travel, cooking and more. This data can be used to train AI models in Human Behavior & Social Interactions, Object & Scene Recognition, Movement & Action Learning, Language & Speech Processing, Cultural & Lifestyle Insights, Temporal & Environmental Changes and "Normal Life" for better context awareness in general. While the footage is not labelled or annotated, it can be upon request.

Type: Video Quantity: 200+ hours Size: 700GB Clarity: 1080p (70%), 4K (30%) This dataset contains high-quality skateboarding footage, capturing a diverse range of tricks, falls, and human movements in real-world environments. It is ideal for AI training in motion recognition, human pose estimation, and action detection across various lighting conditions and perspectives. The footage includes different skaters, locations, and dynamic movements, making it valuable for machine learning applications in sports analytics, robotics, and video understanding. Perfect for AI companies developing computer vision models that require realistic human motion data.

Excel/CSV format, includes restaurant name, URL, address, phone, menu item, menu item description (where available), price for each many item.