Dr. Mori, who specializes in speech and language communication, reported on the progress of the Acoustic Group’s research and discussed the future development of the research.
After confirming the recording conditions required for validation, we discussed what we could do based on the current 10 years(!) worth of voice data we have. We discussed what we can do based on the current recording conditions of the 10 years(!) of voice data we have. There are many variables (=features) that represent the characteristics of speech, and we need to consider them from various perspectives. I myself am still inexperienced in acoustic analysis and have not fully grasped the intuition of analysis, but he gave me advice on how to steadily accumulate findings.
Regarding the future of acoustic research, he pointed out that the problem we are trying to solve, predicting mental health conditions based on acoustic features, could be a new stimulus for the somewhat mature field of speech recognition. We then discussed what we can do for the development of interdisciplinary mental health research, including engineering and psychology. For this purpose, what we, the Acoustics Group, can do is to “visualize” acoustic features using engineering techniques accumulated so far, have psychological researchers and students and graduate students who aspire to become such researchers look at them, identify psychological problems through collaboration, and develop technologies to solve such problems in collaboration with engineering. On the way home after the event, as I was soaking in the afterglow of the event, I was imagining that it might be a bridging work, such as collaborating with engineering to develop technologies to solve such problems.
It is very difficult to extract meaningful information from the audio, but it is a great feeling when you can “see” something that is invisible. It was a day that made me want to increase my weapons (=analysis techniques) and friends so that I can discover more and more.