Multiple Biometrics Use TRIZ Cases
- In-Memory Biometric Comparison Using Parallel CMOS Logic Blocks
Parallel CMOS logic under a memory array compares biometric subsets to stored templates, speeding authentication while improving reliability.
- Multimodal Face Liveness Detection for Deepfake Spoofing
Combining image and video cues, the BiMoTranS model improves face liveness detection against Deepfake and other biometric spoofing attacks.
- Automated Time Recording Using Vision-Based Multi-Employee Recognition
Computer vision and CNN models identify multiple employees and time actions at once, cutting clock-in queues and processing delays.
- Automated Time Recording Using ML-Based Multi-Employee Recognition
Image-based recognition captures multiple employees at once to automate clock-in events, cut wait times, and raise time recording throughput.
- Continuous Biometric Authentication Using Normal-Posture Feature Linking
Links high-accuracy features captured in normal posture with continuous tracking features to keep person identification reliable during movement.
- Multi-Modal Identity Recognition for Occlusion and Large-Angle Scenes
Fusing face and body features improves identity recognition accuracy in complex scenes with occlusion, large angles, and time-based changes.