Expansion in speech time can restore comprehension in a simultaneously speaking bilingual robot

Area of Science:

• Robotics
• Speech Processing
• Human-Computer Interaction

Background:

• Background noise negatively impacts speech comprehension, especially when it contains speech-like properties.
• Simultaneously speaking bilingual robots face challenges due to speech-on-speech interference.
• Improving user experience in such robots requires addressing speech comprehension difficulties.

Purpose of the Study:

• To investigate the effect of time expansion on speech comprehension in a multi-talker environment.
• To enhance the performance of a social robot capable of simultaneous multilingual speech.
• To improve user experience with bilingual robots by mitigating background speech interference.

Main Methods:

• Implemented sentence recognition, speech comprehension, and subjective evaluation tasks.
• Investigated the impact of reduced speech rate and increased pause duration (time expansion) on speech.
• Compared performance in time-expanded multi-talker speech versus standard-speed single-talker speech.

Main Results:

• Time expansion, characterized by reduced speech rate and longer pauses, significantly improved sentence recognition and speech comprehension.
• Participants performed better in time-expanded multi-talker speech compared to standard-speed single-talker speech.
• The positive effect was specific to multi-talker scenarios, as time expansion in single-talker speech did not yield similar improvements.

Conclusions:

• Time expansion of speech, including background speech, facilitates comprehension in simultaneously speaking bilingual robots.
• The presence of background speech can be beneficial if presented in a time-expanded manner.
• Findings suggest a novel approach to designing more effective and user-friendly multilingual social robots.