Robot Audio Direction Control Around Obstacle-Reflected Sound
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Service robots face challenges in accurately detecting the direction of sound sources due to distortion of sound signals caused by obstacles, which affects their ability to transmit and process speech data effectively.
Innovation Solution
A robot system equipped with a vision recognition-based camera, inertial measurement sensors, and multiple microphones, which detects movement, identifies obstacles, estimates their occupation area, and processes sound signals to minimize distortion, thereby improving sound source direction estimation and beamforming performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sound signals are received through multiple microphones in the presence of obstacles, then the robot can detect sound sources, but the sound signals become distorted due to reflection from obstacles
Solution Approach 1:
The patent segments the sound signal processing by separating direct sound paths from reflected sound paths. The controller identifies and processes sound signals based on their origin direction, effectively segmenting the problematic reflected signals from useful direct signals. This allows the robot to focus on sound sources while filtering out obstacle-reflected distortions.
Solution Approach 2:
The patent introduces vision recognition (camera) as an intermediary to detect obstacles and estimate their occupation areas in space. This intermediary information about obstacle locations is then used to identify and filter sound signals that are reflected from these obstacles, resolving the distortion problem without directly modifying the acoustic environment.
2Productivity
If the robot transmits speech data through a network to a server, then speech recognition can be performed, but the robot cannot overcome the problem of distorted sound signals due to obstacles
Solution Approach 1:
The patent performs preliminary action by detecting obstacles and estimating their occupation areas before processing sound signals. The controller proactively identifies the spatial regions occupied by obstacles and uses this information to pre-filter sound signals that would be reflected from these areas, preventing distorted signals from being transmitted to the server for recognition.
3Measurement precision
If the robot estimates sound source direction using multiple microphones, then direction detection can be performed, but the estimation accuracy decreases when obstacles are present
Solution Approach 1:
The patent uses vision recognition as an intermediary to provide obstacle location information that mediates between the microphones and the sound source direction estimation. By combining visual obstacle detection with acoustic signal processing, the system can accurately estimate sound source directions even in the presence of obstacles that would otherwise cause signal distortion and information loss.
Data Source
AI summary
A robot and operation method is disclosed. The robot according to the present disclosure may include a sensor, a microphone, and a controller. The robot may execute an artificial intelligence (AI) algorithm and/or a machine learning algorithm, and may communicate with other electronic devices in a 5G communication environment. An embodiment may include detecting a movement of the robot to a location; detecting an obstacle within a predetermined range from the robot; estimating an occupation area of the obstacle in space; and identifying a sound signal received from the estimated occupation area of the obstacle from among a plurality of sound signals received by a plurality of microphones of the robot at the location.