Based on depth learning algorithm, Firefly face recognition system can off-line face detection, feature extraction and similarity comparison, and open the SDK to provide a complete API.
Open the Ubuntu-based SDK
Realize offline single face recognition
The extreme speed completes face detection and feature matching
Face recognition accuracy
Recognition of many people at the same time
Face recognition performance improved
Open SDK to provide a complete API, achieve faster independent product features
Stand-Alone Work, without remote server, the image inspection, extraction, comparison work
Based on ARM NEON | Mali GPU | Perfblas | CNN Smart acceleration with 280 +
Complete face detection, extraction and comparison <0.5s, and face recognition accuracy > 99%
1:1 face comparison, 1:N face recognition, face location, face authentication
Support USB camera, image file face data acquisition
Support for multiple ways of data exchange, local and remote database access
Depth of optimization Ubuntu , support for proprietary face recognition acceleration framework
Face recognition technology is widely used in video surveillance, import and export control, access control, identity verification and other important security areas.
Integration of high-precision face recognition technology, improve the tax service hall integrated service management level, to achieve integrated management.
Face recognition technology to help banks and other financial institutions to complete the real-name system verification, identity real-time verification, self-help ID card verification.
In the admission examination using face recognition to achieve a document security, standard consistent, evidence to save.
Face recognition technology can be used to store, business building monitoring, gates and other security facilities, and to achieve customer traffic monitoring.