Abstract: A “running range normalization” method includes computing running estimates of the range of values of features useful for voice activity detection (VAD) and normalizing the features by mapping them to a desired range. Running range normalization includes computation of running estimates of the minimum and maximum values of VAD features and normalizing the feature values by mapping the original range to a desired range. Smoothing coefficients are optionally selected to directionally bias a rate of change of at least one of the running estimates of the minimum and maximum values. The normalized VAD feature parameters are used to train a machine learning algorithm to detect voice activity and to use the trained machine learning algorithm to isolate or enhance the speech component of the audio data.

Abstract: A “running range normalization” method includes computing running estimates of the range of values of features useful for voice activity detection (VAD) and normalizing the features by mapping them to a desired range. Running range normalization includes computation of running estimates of the minimum and maximum values of VAD features and normalizing the feature values by mapping the original range to a desired range. Smoothing coefficients are optionally selected to directionally bias a rate of change of at least one of the running estimates of the minimum and maximum values. The normalized VAD feature parameters are used to train a machine learning algorithm to detect voice activity and to use the trained machine learning algorithm to isolate or enhance the speech component of the audio data.

Abstract: A “running range normalization” method includes computing running estimates of the range of values of features useful for voice activity detection (VAD) and normalizing the features by mapping them to a desired range. Running range normalization includes computation of running estimates of the minimum and maximum values of VAD features and normalizing the feature values by mapping the original range to a desired range. Smoothing coefficients are optionally selected to directionally bias a rate of change of at least one of the running estimates of the minimum and maximum values. The normalized VAD feature parameters are used to train a machine learning algorithm to detect voice activity and to use the trained machine learning algorithm to isolate or enhance the speech component of the audio data.

Abstract: The overall loudness of an audio track is calculated by combining a number of weighted loudness measures for segments of the audio track, where the weight applied to each individual loudness measure is a function of the loudness measure. By comparing the original overall loudness measure to a desired overall loudness measure, a gain can be determined that will adjust the loudness level to the desired value. Also disclosed is a dynamic compression method that analyzes the dynamic characteristics of an audio track and determines appropriate compressor parameters. Additionally, the loudness of a post-compressor audio track can be estimated for any given compressor parameters, thus permitting post-compression loudness matching to be done even if the compression is performed in real-time.

Abstract: A cutting tool assembly for machining a turbine component includes a substantially solid toolblock body, a blade sub-assembly, a cutting insert, and a lead screw. The toolblock body includes a centerline axis. The blade sub-assembly is slidably retained adjacent said toolblock body and a cutting insert is attached to the blade sub-assembly. The lead screw coupled adjacent to the toolblock body such that the lead screw is substantially perpendicular to the centerline axis. The lead screw is rotatably coupled to the blade sub-assembly such that rotation of the lead screw extends the blade sub-assembly from the toolblock body.

Abstract: A cutting tool assembly for machining a turbine component includes a substantially solid toolblock body, a blade sub-assembly, a cutting insert, and a lead screw. The toolblock body includes a centerline axis. The blade sub-assembly is slidably retained adjacent said toolblock body and a cutting insert is attached to the blade sub-assembly. The lead screw coupled adjacent to the toolblock body such that the lead screw is substantially perpendicular to the centerline axis. The lead screw is rotatably coupled to the blade sub-assembly such that rotation of the lead screw extends the blade sub-assembly from the toolblock body.