Noise reduction improves memory for target language speech in competing native but not foreign language speech

Area of Science:

• Audiology
• Cognitive Psychology
• Speech Processing

Background:

• Hearing aids often struggle to isolate target speech in noisy environments.
• Noise reduction (NR) algorithms aim to improve speech intelligibility and listening comfort.
• Previous research suggests NR benefits individuals with high working memory capacity.

Purpose of the Study:

• To evaluate the effectiveness of NR for individuals with low working memory capacity.
• To determine how NR affects speech recall when the competing speech is in a non-native language.
• To assess the impact of NR on memory for native speech amidst competing native or non-native speech.

Main Methods:

• A sentence-final word identification and recall (SWIR) test was used with 26 hearing aid users.
• Participants recalled Swedish target speech presented with competing Swedish or Cantonese speech, with or without NR.
• Working memory capacity was assessed using a reading span (RS) test.

Main Results:

• Recall performance correlated with working memory capacity (RS scores).
• The benefit of NR on recall was independent of working memory capacity.
• Native language competing speech disrupted recall more than foreign language speech.
• NR improved recall in the presence of native competing speech but not foreign speech.

Conclusions:

• Hearing aid noise reduction enhances memory for speech in competing sound environments.
• NR mitigates the disruptive effect of native language babble on speech memory.
• The positive impact of NR on speech memory is not limited by working memory capacity.