H92-1037 , we are mainly concerned with spectral normalization . For each frame x a , we want
H89-2044 major differences between various spectral normalization algorithms are primarily concerned
H89-2044 While Ward was especially and spectral normalization , and each of the two microphones
H89-2044 Sphinx system , and we describe the spectral normalization and spectral subtraction operations
H92-1055 either spectral subtraction or spectral normalization . The horizontal dotted lines
H89-2044 procedures we used to achieve spectral normalization and spectral subtraction in the
H89-2044 estimated . The most effective form of spectral normalization that we have considered so far
H89-2044 also considered other types of spectral normalization in the cepstral domain , including
H89-2044 Spectral Normalization The goal of spectral normalization is to compensate for distortions
H89-2044 baseline conditions , or with spectral normalization and spectral subtraction . In
H89-2044 problems : spectral subtraction and spectral normalization . In spectral subtraction one
H89-2044 isolation than the simple static spectral normalization described above , some of them
H92-1056 Therefore , some type of additional spectral normalization is required to compensate for
H92-1055 processing , spectral subtraction , spectral normalization , and the CDCN algorithm . SPHINX
H89-2044 Subtraction is more complex than spectral normalization , both because it can not be
H89-2044 that in each case , the use of spectral normalization and subtraction provides increasing
H89-2044 reduction in relative error rate when spectral normalization was added to spectral subtraction
H92-1055 processing , spectral subtrac - tion , spectral normalization , and the CDCN algorithm . SPmNX-II
H89-2044 signal processing procedures . Spectral Normalization The goal of spectral normalization
H89-2044 ; Berouti et al. , 1979 ) . In spectral normalization ( sometimes referred to as "
hide detail