Abstract: The information processing apparatus (2000) computes a first score representing a degree of similarity between the input sound data (10) and the registrant sound data (22) of the registrant (20). The information processing apparatus (2000) obtains a plurality of pieces of segmented sound data (12) by segmenting the input sound data (10) in the time direction. The information processing apparatus (2000) computes, for each piece of segmented sound data piece (12), a second score representing the degree of similarity between the segmented sound data (12) and the registrant sound data (22). The information processing apparatus 2000 makes first determination to determine whether a number of speakers of sound included in the input sound data (10) is one or multiple, using at least the second score.

Abstract: Disclosed herein are systems and methods for identifying the source of a signal via channel equalization using characteristics of the signal. A system receives a signal, then measures a frequency response of the signal by performing a spectral analysis over the entire signal. The system computes the average amplitude over a subset of time samples from the spectral analysis for each represented frequency and compares the set of averaged amplitudes to a stored set of averaged amplitudes to produce equalization coefficients. Applying the equalization coefficients to the frequency response yields an equalized frequency response, which is compared to a stored frequency response using a classifier to determine a match. Alternately, the system applies the equalization coefficients to the stored frequency response yielding an equalized stored frequency response. The method can recognize speakers, vehicles, electromagnetic signals, sonar signals, optical signals, videos, etc.

Abstract: Embodiments of a progressive feature selection method that selects features in multiple rounds are described. In one embodiment, the progressive feature selection method splits the feature space into tractable sub-spaces such that a feature selection algorithm can be performed on each sub-space. In a merge-split operation, the subset of features that the feature selection algorithm selects from the different sub-spaces are merged into subsequent sets of features. Instead of re-generating the mapping table for each subsequent set from scratch, a new mapping table from the previous round's tables is created by collecting those entries that correspond to the selected features. The feature selection method is then performed again on each of the subsequent feature sets and new features are selected from each of these feature sets. This feature selection-merge-split process is repeated on successively smaller numbers of feature sets until a single final set of features is selected.

Abstract: Embodiments of a feature generation system and process for use in machine learning applications utilizing statistical modeling systems are described. In one embodiment, the feature generation process generates large feature spaces by combining features using logical, arithmetic and/or functional operations. A first set of features in an initial feature space are defined. Some or all of the first set of features are processed using one or more arithmetic, logic, user-defined combinatorial processes, or combinations thereof, to produce additional features. The additional features and at least some of the first set of features are combined to produce an expanded feature space. The expanded feature space is processed through a feature selection and optimization process to produce a model in a statistical modeling system.

Abstract: A system and method are provided to authenticate a voice in a frequency domain. A voice in the time domain is transformed to a signal in the frequency domain. The first harmonic is set to a predetermined frequency and the other harmonic components are equalized. Similarly, the amplitude of the first harmonic is set to a predetermined amplitude, and the harmonic components are also equalized. The voice signal is then filtered. The amplitudes of each of the harmonic components are then digitized into bits to form at least part of a voice ID. In another system and method, a voice is authenticated in a time domain. The initial rise time, initial fall time, second rise time, second fall time and final oscillation time are digitized into bits to form at least part of a voice ID. The voice IDs are used to authenticate a user's voice.

Abstract: Disclosed is a method for speech speaker recognition of a speech speaker recognition apparatus, the method including detecting effective speech data from input speech; extracting an acoustic feature from the speech data; generating an acoustic feature transformation matrix from the speech data according to each of Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA), mixing each of the acoustic feature transformation matrixes to construct a hybrid acoustic feature transformation matrix, and multiplying the matrix representing the acoustic feature with the hybrid acoustic feature transformation matrix to generate a final feature vector; and generating a speaker model from the final feature vector, comparing a pre-stored universal speaker model with the generated speaker model to identify the speaker, and verifying the identified speaker.