Based on face classification, image quality analysis, and face clustering techniques trained by deep learning, Tetras provide the face clustering product to acquire the faces and capture facial features from the source materials (pictures and videos). In addition, the machine can automatically select the appropriate group of people for classification and combination with the optimal decision-making process of the stochastic dynamic system. After face clustering, the algorithm can comprehensively criteria the face quality, picture sharpness, and aesthetic appearance in the new album, with various dimensions. After that, it will automatically select the optimal face photo as the album cover, which is a user-friendly process. Tetras face clustering product can support verification, extraction, and clustering of different faces and face features in plenty of photos with an accuracy rate of over 99%, without any platform, frame rate, or resolution restrictions. The product supports both Android and iOS operating systems, cloud services, and all mainstream platforms. Tetras already have a healthy long-term business relationship with numerous smartphone brands
Supports Android and Linux platforms; and the CPU version supports a full range of Qualcomm and MTK product models.
Tetras.AI's facial clustering products have been leading the AI visual market for a long time. Its algorithm can process face verification, feature extraction, and face clustering of thousands of different faces, with ultra-high accuracy (99%+) and recall rate (90%+). It supports the application scenarios of different age ranges, even though the younger generation, like children, also holds a high recall rate. At the same time, it supports large-scale crowd application scenarios and face clustering of people worldwide. We actively respond to market changes and satisfy varying customer needs.
By detecting facial information, capturing and comparing facial features, Tetras.AI's face clustering product groups similar features for classification to accurately identify multiple facial attribute categories and provides analytical data from the perspectives of picture quality and aesthetics.
Determines the coordinate position of a face in an image, extracts it, and outputs a face frame
Performs feature comparison after capturing and recognizing the extracted face features
Accurately identifies multiple attribute categories, such as gender, smile, age, etc
Analyzes image quality from multiple dimensions, such as color, noise, blur, fog sense, exposure, and other dimensions
Performs aesthetic analysis of images from the dimensions of facial expressions, facial movements (e.g., eyes open and close), and face image quality (e.g., exposure, sharpness comparison, angle, occlusion and color)
For mobile phone system album, Cloud album facial clustering features, including album classification, facial attributes classification, video cover, album cover generation, people search function.