Objective: Backward-facing directional processing (Back-DIR) is an algorithm that employs an anti-cardioid directivity pattern to enhance speech arriving from behind the listener. An experiment that was originally designed to evaluate Back-DIR, together with its follow-up experiment, are reported to illustrate how hearing-aid users' voices could affect directional benefit.
Design: Speech recognition performance was measured in a speech-180°/noise-0° configuration, with aids programmed to Back-DIR enabled or omnidirectional processing. In the original experiment, the conventional hearing-in-noise test (HINT) was used, wherein listeners repeated heard sentences. In the follow-up experiment, a modified HINT was used, wherein a carrier phrase was presented before each sentence.
Study Sample: Fifteen adults with sensorineural hearing loss participated in both experiments.
Results: Significant Back-DIR benefit (relative to omnidirectional processing) was observed in the follow-up experiment, while not in the original experiment.
Conclusions: In the original experiment, hearing aids were affected by listeners' voices such that Back-DIR was not always activated when the target speech was presented. In the follow-up experiment, listeners' voice effects were eliminated by the carrier phrase activating Back-DIR before the sentences were presented. The results suggest that the effect of hearing-aid technologies is highly dependent on the characteristics of listening conditions.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.3109/14992027.2013.802381 | DOI Listing |
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!