Liveness Detection via Random Action Sequences and Multi-Sensor Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current human face recognition systems lack effective liveness detection methods to prevent attacks using images, videos, 3D models, or face masks, often relying on specialized hardware.
Innovation Solution
A liveness detection method using a random action sequence generated by combining image and non-image sensor data, where the system determines if randomly sent actions are executed by a living body, incorporating image and non-image sensor information to verify liveness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If specialized hardware devices (infrared camera, depth camera) are used for liveness detection, then detection reliability is improved, but device complexity and cost increase
Solution Approach 1:
The patent replaces specialized hardware devices (infrared cameras, depth cameras) with a combination of standard image sensor and non-image sensors (accelerometer, gyroscope, light sensor). The mechanical/optical detection system is substituted with a multi-sensor fusion system using conventional components, thereby reducing device complexity while maintaining detection reliability through software-based liveness detection algorithms.
Solution Approach 2:
The patent makes the mobile device's existing sensors serve multiple functions: the image sensor captures both authentication images and liveness detection data, while non-image sensors (accelerometer, gyroscope, light sensor) that originally served other purposes are repurposed to detect action instructions and determine liveness. This multi-functionality eliminates the need for dedicated liveness detection hardware.
2Reliability
If simple still picture attacks are prevented, then basic security is improved, but vulnerability to advanced attacks (video, 3D model, face mask) remains
Solution Approach 1:
The patent transitions from static liveness detection (preventing still picture attacks) to dynamic liveness detection by introducing random action instructions that require the user to perform sequential actions. The system detects whether the user is alive by verifying they can execute dynamic, time-dependent actions, which prevents attacks using static images, pre-recorded videos, 3D models, and face masks.
Solution Approach 2:
The patent implements preliminary anti-action by requiring the user to complete a sequence of random actions before authentication is granted. This preliminary dynamic verification must occur in real-time, preventing attackers from using pre-prepared attack materials. The system sends multiple random action instructions in sequence, and the user must execute them in the correct order, creating a time-dependent security barrier.
3Measurement precision
If random action sequence with multiple sensors is used, then liveness detection accuracy is improved, but processing complexity increases
Solution Approach 1:
The patent segments the liveness detection process into distinct stages: (1) sending random action instructions, (2) capturing images and sensor data, (3) determining whether actions are executed correctly, and (4) making the final liveness determination. Each stage uses specific sensors and processing methods, allowing the system to manage complexity through structured segmentation while maintaining high detection accuracy.
Data Source
AI summary
There are provided a liveness detection method and device. The liveness detection method comprises: generating a random action instruction sequence including at least one random action instruction; sequentially sending a random action instruction in the random action instruction sequence; and determining whether the sequentially sent random action instruction in the random action instruction sequence is sequentially executed by a living body based on detection information of at least two sensors, wherein the at least two sensors comprise an image sensor and at least one non-image sensor; and determining that the liveness detection is succeeded if the sequentially sent random action instruction in the random action instruction sequence is sequentially executed by the living body. Accuracy of liveness detection can be improved by adopting the random action sequence and by combining images captured by the image sensor and information detected by the non-image sensor.