To tackle this problem, researchers have designed a new algorithm BOSSA. It tackles the most remarkable “cocktail party problem” by sorting out individual voices in crowded spaces. A recent round of lab tests on BOSSA, inspired by barn owls, by a team of scientists created an exciting buzz. These incredible gregarious avian navigators have developed advanced spatial sensing practices, truly navigating through echolocation and enabling them to find prey by listening. The results of this study were released April 22 in the open-access journal Communications Engineering.
BOSSA is especially concerned with the spatial disparity between two sounds. This method mimics the eco-location that barn owls use to navigate direction and distance as they hunt. This algorithm is less computationally intensive than most state-of-the-art sound processing algorithms. It’s potentially a pretty useful tool for people who are hearing impaired.
In a tightly controlled experiment, scientists equipped people with hearing loss with custom-built headphones. These headphones faithfully reproduced sounds that made it feel like five people were talking to you simultaneously from all sides. The riders’ challenge was to keep up with sentences read by any of the five 10th planet instructors. This effect was particularly vivid in that when the target speaker was within 30 degrees of the listener, participants were able to understand more words even at the quieter volume. Compared to the major improvement seen in either direction when using BOSSA.
When tested in actual listening conditions, BOSSA produced statistically significant improvements over a conventional algorithm and unassisted listening. The findings demonstrated that BOSSA has the potential to significantly enhance speech intelligibility for people with hearing impairment. This amplification only takes place in specific spatial arrangements.
Michael Stone, an audiology researcher at the University of Manchester in the U.K., praised the potential of BOSSA. As he explained, the technology lacks the ability to recreate the nuanced sound echo and reverberation effects present in actual indoor settings.
“You can’t do that with this algorithm.” – Michael Stone
Fan-Gang Zeng, one of the researchers involved in the study, recognized that while BOSSA is a promising advancement, it may require further refinement. Zeng described it as “a new flashlight that has a tighter beam that’s more selective,” illustrating how BOSSA enhances focus on specific sounds while filtering out background noise.
The present study tested BOSSA with only four trained participants. That’s a good thing because it indicates that we need further research to replicate its effectiveness in a more generalized population. As we study BOSSA in more practical settings, its possible uses will continue to expand. This expansion opens up new possibilities to provide better listening experiences for people with hearing difficulties.
