Variable-Bandwidth Noise Filtering Mechanism for the Hearing Aid System

Enhancement of an hearing aid (HA) is the most needed task for people with hearing problems. The conventional filtering module doesn't provide a better filtering report and needs much duration for the filtering function. So, the current research article has discovered a novel CatBoost-based Variable-Bandwidth filter (CbVBwF) to enrich the noise removal process and filter the noise features in the trained audio signal. Primarily, the audio signals are trained using the training features of CatBoost. Consequently, the boosting parameters in the novel CbVBwF offered a better signal filtering function within a short duration, which has helped to earn less delay score and matching flaw rate. Henceforth, the designed framework is evaluated in the MATLAB platform, and the chief metrics were validated with other past studies. The novel CbVBwF scored less matching flaw score of 1.2 dB than the compared techniques.