Robot Gesture Noise Reduction via Paired Profiles

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Robot-human communication is hindered by mechanical noise generated during robot gestures, which interferes with speech recognition and reduces the signal-to-noise ratio, making it difficult for robots to accurately interpret user utterances.

Innovation Solution

A method and apparatus that utilize a gesture library to pair predetermined robot gestures with noise profiles, allowing for the removal of mechanical noise from incoming audio, thereby enhancing speech recognition by improving the signal-to-noise ratio.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If the robot performs gestures to enhance natural communication, then the communication naturalness is improved, but mechanical noise is generated that interferes with speech recognition

Engineering Contradiction:

Improvecommunication naturalnessVSAvoidmechanical noise interference

Core Design Contradiction:

Adaptability or versatilityVSObject-affected harmful factors

Solution Approach 1:

The audio signal is segmented into gesture noise components and speech components. The system separates the mechanical noise generated during gestures from the user's speech by identifying and isolating the noise portion, allowing independent processing of each component to preserve speech quality while removing interference.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The gesture noise is extracted from the mixed audio signal using acceleration sensors that detect robot motion. The extracted noise profile is then removed from the incoming audio, leaving only the clean speech signal for recognition, thereby eliminating the harmful noise while preserving the beneficial gesture communication.

Inventive Principle:

Principle #2Taking out (Extraction)

2Measurement precision

If the robot uses speech recognition to interpret user utterances, then the ability to understand user input is improved, but mechanical noise from gestures reduces the signal-to-noise ratio and decreases recognition accuracy

Engineering Contradiction:

Improvespeech recognition accuracyVSAvoidsignal-to-noise ratio degradation

Core Design Contradiction:

Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The system uses acceleration sensors to continuously monitor robot motion and provides feedback about gesture execution. This feedback loop allows the speech recognition system to dynamically adjust by subtracting the predicted gesture noise from the audio input, maintaining high recognition accuracy even during active gestures.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system treats the gesture noise and speech signal asymmetrically by identifying and removing only the noise portion while preserving the speech content. The acceleration data is used to create a noise profile that is specifically targeted for removal, leaving the speech signal intact and improving the effective signal-to-noise ratio for recognition.

Inventive Principle:

Principle #4Asymmetry

Data Source

PatentEP3966817B1Noise reduction in robot human communication

Publication Date: 2023.12.13 MICROSOFT TECHNOLOGY LICENSING LLC

Figure 1
Figure 2
Figure 3A~3B

AI summary

Noise reduction in a robot system includes the use of a gesture library that pairs noise profiles with gestures that can be performed by the robot. A gesture to be performed by the robot is obtained, and the robot performs the gesture. The robot's performance of the gesture creates noise, and when a user speaks to the robot while the robot performs a gesture, incoming audio includes both user audio and robot noise. A noise profile associated with the gesture is retrieved from the gesture library and is applied to remove the robot noise from the incoming audio.