Endpoint detection based on MDL using subband speech satisfied auditory model

An adaptive endpoint detection algorithm based on band energy and adaptive smoothing algorithm has been described. This algorithm utilized the capability of adaptive smoothing algorithm that intensifies the discontinuity between local areas. The selection of the gradient threshold utilizes the MDL (minimal description length) criterion. We selected the band energy features because of its usefulness in detecting high-energy regions (in the incoming signal) and making the distinction between speech and noise. Heuristic "Edge-focusing" is used to endpoint detection to save the consuming-time in iterative.