Active headrest combined with a depth camera-based ear-positioning system

Yuteng Liu,Haowen Li,Haishan Zou,Jing Lu,Zhibin Lin
2023-12-25
Abstract:Active headrests can reduce low-frequency noise around ears based on active noise control (ANC) system. Both the control system using fixed control filters and the remote microphone-based adaptive control system provide good noise reduction performance when the head is in the original position. However, their performance degrades significantly when the head is in motion. In this paper, a human ear-positioning system based on the depth camera is introduced to address this problem. The system uses RTMpose model to estimate the two-dimensional (2D) positions of the ears in the color frame, and then derives the corresponding three-dimensional (3D) coordinates in the depth frame with a depth camera. Experimental results show that the ear-positioning system can effectively track the movement of ears, and the broadband noise reduction performance of the active headrest combined with the system is significantly improved when the human head is translating or rotating.
Computer Vision and Pattern Recognition,Image and Video Processing
What problem does this paper attempt to address?
The paper aims to address the issue of decreased low-frequency noise control effectiveness in active headrests during head movements. Specifically, existing active headrest systems can effectively reduce noise when the head position is fixed, but their performance significantly declines when the head moves. To this end, the paper proposes an Ear-Positioning System (EP system) that combines a depth camera to improve the noise reduction performance of active headrests by tracking the position of the ears in real-time. The specific implementation methods include: 1. **Ear Positioning**: Using the RTMpose model to estimate the 2D position of the ears in color images and obtaining the corresponding 3D coordinates through a depth camera. 2. **Occlusion Handling**: When one ear is occluded, a post-processing algorithm based on symmetry is proposed to estimate the 3D position of the occluded ear. 3. **Experimental Validation**: The system's positioning accuracy and noise reduction effect when combined with active headrests were validated through experiments. The results show that the system can significantly improve the control effect of broadband noise during head translation or rotation.