Robust spectro-temporal speech features with model-based distribution equalization | IEEE Conference Publication | IEEE Xplore