研究目的
Investigating the capability of temporal deep neural networks to interpret natural human kinematics for active biometric authentication with mobile inertial sensors.
研究成果
The study concludes that human kinematics convey important information about user identity and can serve as a valuable component of multi-modal authentication systems. The proposed model also shows promise for application in a visual context.
研究不足
The study is limited by the challenges of collecting data from a practical and legal perspective, the noisy nature of inertial data, and the need for efficient learning and incorporation of features into a biometrics setting with limited resources.
1：Experimental Design and Method Selection:
The study involves comparing several neural architectures for efficient learning of temporal multi-modal data representations, proposing an optimized shift-invariant dense convolutional mechanism, and incorporating the discriminatively trained dynamic features in a probabilistic generative framework.
2：Sample Selection and Data Sources:
A first-of-its-kind data set of human movements was passively collected by 1500 volunteers using their smartphones daily over several months.
3：List of Experimental Equipment and Materials:
Smartphones with inertial sensors (accelerometer and gyroscope) were used for data collection.
4：Experimental Procedures and Operational Workflow:
Data was collected passively from volunteers using their smartphones, focusing on natural human kinematics. The data was then used to train and compare several neural architectures.
5：Data Analysis Methods:
The study employs probabilistic generative frameworks and discriminative training for dynamic feature extraction and analysis.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
