ve a ing ous i ven De e ion via ee ua y ns i ed he Ho is i ased e esen a ion Me hod

ABSTRACT A novel dictionary learning approach that utilizes Mel-scale frequency warping in detecting overlapped acoustic events is proposed. The study explored several dictionary learning schemes for improved performance of overlapping acoustic event detection. The structure of NMF for calculating gains of each event was utilized for including in overlapped signal for its low computational load. In this paper, we propose a method of frequency warping for better sound representation, and apply dictionary learning by a holistic-based representation, namely nonnegative K-SVD (NK-SVD) in o rder to resolve a basis sharing problem raised by part-based representations. We confirm that the proposed method of Mel-scale with NK-SVD delivered significantly better results than the conventional methods.