Overview
Visual search differs significantly from search engines based on text. It is a new level of information analysis. It requires indexing of media files or live video streams, automatic extraction of important information in compact form, which later on can be used for search directly as well as in combination with traditional text metadata. Our computer vision and machine learning technology provides scalable solution, unrelated to current model of relying on human interaction to tag, describe, and classify images as more photos, videos, and other forms of rich media find their way online.
We provide new automatic metadata extraction and search possibilities based on proprietary, unique algorithms that overperforms any other methods. Our advanced face recognition, employed independently or in combination with speech and text recognition provides unbeatable accuracy, speed and scale of search results for images and video data.
In addition to core recognition and analysis processes we employ other face-related measurements like detection of hairstyle, hair color and skin color, facial hair detection, glasses detection, shape analysis of facial features to automatically extract additional metadata.
Visual search differs significantly from search engines based on text. It is a new level of information analysis. It requires indexing of media files or live video streams, automatic extraction of important information in compact form, which later on can be used for search directly as well as in combination with traditional text metadata. Our computer vision and machine learning technology provides scalable solution, unrelated to current model of relying on human interaction to tag, describe, and classify images as more photos, videos, and other forms of rich media find their way online.
We provide new automatic metadata extraction and search possibilities based on proprietary, unique algorithms that overperforms any other methods. Our advanced face recognition, employed independently or in combination with speech and text recognition provides unbeatable accuracy, speed and scale of search results for images and video data.
In addition to core recognition and analysis processes we employ other face-related measurements like detection of hairstyle, hair color and skin color, facial hair detection, glasses detection, shape analysis of facial features to automatically extract additional metadata.
Technology and Company highlights
- Years of research focused on facial recognition;
- Years of experience creating customized face recognition solutions for our clients, including interactive online applications and web services
- Robust, working server-side software for face detection, facial points detection, faces comparison and persons identification in images, video files and live video streams
- Advanced face appearance analysis, biometric measurements
- High scalability, can process thousands of images per server or multiple video streams simultanously (2+ streams per server);
- Able to identify known persons or extract unknown persons appearances for later identification;