Human-Mounted Biosensors for 3D Facial Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional facial landmark tracking and 3D reconstruction methods require users to be confined to a specific location with constrained recording conditions, such as no occlusions and good lighting, which limits their application in scenarios involving human motion.
Innovation Solution
The development of a human-mounted biosensor system that uses electromyography (EMG) and electrooculography (EOG) information to perform biosensor-based user authentication and identification, employing a multi-input convolutional neural network (CNN) and Long Short-Term Memory (LSTM) to extract spatial and temporal representations, which are then processed by a machine-learning classifier for authentication or identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional camera-based facial landmark tracking is used, then 3D reconstruction accuracy is improved, but user mobility and operational flexibility deteriorate due to constrained recording conditions
Solution Approach 1:
The patent replaces camera-based optical measurement systems with a biosensor system that uses electromyography (EMG) and electrooculography (EOG) electrodes mounted on the user's head. This substitution eliminates the need for external cameras and controlled lighting conditions, allowing accurate 3D facial reconstruction while maintaining user mobility and comfort during natural movements.
Solution Approach 2:
The patent introduces biosensors (EMG and EOG electrodes) as intermediary devices that directly contact the user's facial muscles and eye regions. These biosensors capture physiological signals that serve as intermediaries between facial movements and the computational model, enabling accurate tracking without requiring line-of-sight cameras or controlled environments.
2Difficulty of detecting and measuring
If camera-based facial tracking is used, then facial expression detection capability is improved, but adaptability to various recording conditions deteriorates
Solution Approach 1:
The patent replaces camera-based visual detection with biosensor-based physiological signal detection. EMG electrodes detect electrical activity from facial muscles, while EOG electrodes detect eye movement potentials. This substitution enables facial expression detection that is independent of lighting conditions, occlusions, and camera angles, significantly improving adaptability to various recording environments.
Solution Approach 2:
The biosensor system captures physiological signals directly from the user's own body (facial muscles and eyes), making the detection process self-contained and independent of external environmental factors. The user's own physiological signals serve as the measurement source, eliminating reliance on external lighting, camera quality, or controlled settings.
3Adaptability or versatility
If biosensors are mounted on the user's head, then adaptability to human motion is improved, but device complexity increases
Solution Approach 1:
The patent implements a multi-functional head-mounted device that integrates both EMG and EOG sensing capabilities within a single biosensor system. This unified approach allows the device to simultaneously capture facial muscle activity and eye movements, providing comprehensive facial expression data while maintaining a relatively simple overall structure compared to multiple separate sensing systems.
Solution Approach 2:
The patent uses the user's head and face as the mounting structure and signal source, eliminating the need for external cameras, lighting equipment, and complex positioning systems. The biosensors directly contact facial surfaces, using the body itself as the support structure, thereby reducing external device complexity while improving adaptability to natural head movements.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This system enables continuous and unobtrusive tracking of facial movements and expressions without visual input, allowing for accurate 3D facial reconstruction and authentication/identification, comparable to state-of-the-art camera-based solutions, with minimal impact on user mobility and comfort.
Implementation Method 1
receiving user-specific biosignal information from electrodes, where the user-specific biosignal information includes electromyography (EMG) information
Implementation Method 2
the user-specific biosignal information includes electromyography (EMG) information and electrooculography (EOG) information
Data Source
AI summary
Provided are methods, apparatus, systems, and computer-readable media for processing biosensor information to produce multi-dimensional images of facial movement, produce multi-dimensional images of expressions of a human user, perform biosensor-based user identification, perform biosensor-based user authentication, and combinations thereof. Other methods, apparatus, systems, and computer-readable media are also disclosed.