Dynamic Beamforming for Mobile Audio Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computers and mobile devices struggle to include multiple individuals in a call from different positions in a room effectively due to limitations in beamforming techniques, which result in background noise and voice muffling, especially under poor signal-to-noise ratio conditions and device movement.
Innovation Solution
A mobile platform with a microphone array and orientation sensors that continuously adjusts beamforming to maintain audio focus on specific sound sources despite device movement, using a combination of accelerometer, gyroscope, and digital compass data to steer the beamforming direction and suppress unwanted noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If beamforming techniques using beam steering algorithms are used to reduce ambient noise and identify talkers, then background noise suppression is improved, but voice muffling and background noise modulation occur under poor signal-to-noise ratio conditions and during fast switches between talkers
Solution Approach 1:
The patent implements dynamic beamforming by continuously updating the direction of arrival estimates and beamforming weights in real-time as the device moves. The system adapts the beamforming parameters dynamically based on current orientation sensor data and acoustic environment, rather than using fixed beam steering algorithms. This allows the system to maintain reliable voice capture across changing conditions without the muffling artifacts that occur with static beam steering approaches.
Solution Approach 2:
The system changes the parameters of beamforming based on detected device movement and orientation. When movement is detected via orientation sensors, the system updates the beamforming direction and focal point parameters to track the new device orientation. This parameter adaptation allows the system to maintain optimal voice capture while suppressing background noise, resolving the contradiction between noise suppression and voice clarity under varying conditions.
2Measurement precision
If directional microphones or microphone arrays with fixed beamforming are used, then audio focus on a single direction is improved, but the system cannot easily include persons at different positions in the room
Solution Approach 1:
The patent makes the beamforming system dynamic by allowing real-time reconfiguration of the beam direction and focal points. When a new talker enters the scene or the device orientation changes, the system dynamically updates the beamforming parameters to include the new speaker while maintaining audio focus. This dynamic adaptability allows the system to switch between capturing different positions in the room, resolving the contradiction between directional precision and multi-position versatility.
Solution Approach 2:
The system achieves multi-functionality by combining fixed beamforming capabilities with dynamic reconfiguration. The microphone array can serve multiple functions: maintaining fixed directional focus when needed, switching to track moving talkers, and adapting to include multiple persons at different positions. This universal approach allows the same hardware to provide both precise audio directionality and flexible multi-position talker inclusion based on situational requirements.
3Ease of operation
If the device is moved during conversation, then user mobility is improved, but direction of arrival identification becomes difficult causing voice muffling and background noise modulation
Solution Approach 1:
The system performs preliminary actions by continuously monitoring orientation sensor data and pre-calculating beamforming adjustments before significant movement affects audio quality. When movement is detected via orientation sensors, the system proactively updates the beamforming direction and focal points to compensate for the device repositioning. This preliminary adjustment maintains accurate direction of arrival identification even during device movement, resolving the contradiction between mobility and identification accuracy.
Solution Approach 2:
The patent implements feedback by continuously monitoring orientation sensor data and using this information to adjust beamforming parameters in real-time. The system creates a feedback loop where movement detection triggers beamforming reconfiguration, which then maintains accurate direction of arrival identification despite device relocation. This feedback mechanism allows the system to maintain reliable audio capture during mobility, resolving the contradiction between ease of movement and directional accuracy.
Data Source
Figure 1A~3
Figure 2A~2C
Figure 4A~4C
AI summary
A mobile platform includes a microphone array and is capable of implementing beamforming to amplify or suppress audio information from a sound source. The sound source is indicated through a user input, such as pointing the mobile platform in the direction of the sound source or through a touch screen display interface. The mobile platform further includes orientation sensors capable of detecting movement of the mobile platform. When the mobile platform moves with respect to the sound source, the beamforming is adjusted based on the data from the orientation sensors so that beamforming is continuously implemented in the direction of the sound source. The audio information from the sound source may be included or suppressed from a telephone or video-telephony conversation. Images or video from a camera may be likewise controlled based on the data from the orientation sensors.