Machine Learning Based Dynamic Band Selection for Splitting Auditory Signals to Reduce Inner Ear Hearing Losses

Quality of hearing has been severely impacted due to signal losses occurs in the human inner ear explicitly region cochlea. Loudness recruitment, degraded frequency selectivity and auditory masking are major outward effects losses. Splitting signals into bands presenting dichotically both ears became a comprehensive solution reduce However, these methods divide input fix number bands, this limits their applicability where have large variations spectral characteristics. To address challenge, we proposed machine learning based intelligent band selection algorithm split dynamically. Proposed analyze speech on characteristics determine optimum required effectively present acoustic cues signal. Further, dynamic splitting efficiently divides for dichotic presentation. method examined subjects from different age groups gender having cochlear impairment. Qualitative quantitative assessment shown significant improvement recognition score with substantial reduction response time.