Abstract: A computer system that processes mixtures of signals, such as speech and noise sources derived from multiple simultaneous microphone recordings, in order to separate them into their underlying sources. A source separation routine optimizes a filter structure by minimizing cross powers of multiple output channels while enforcing geometric constraints on the filter response. The geometric constraints enforce desired responses for given locations of the underlying sources, based on the assumption that the sources are localized in space.

Abstract: An iterative approach to vector median filtering wherein the resulting median vector need not be a member of the original data set. The iterative vector median filtering allows for fast convergence for complex computations and an output which is approximate to the mean, particularly for small data sets. In addition, a method and system for registering and matching 2.5 normal maps is provided. Registration of two maps is performed by optimally aligning their normals through 2-D warping in the image plane in conjunction with a 3-D rotation of the normals. Once aligned, the average dot-product serves as a matching metric for automatic target recognition (ATR).

Abstract: An iterative approach to vector median filtering wherein the resulting median vector need not be a member of the original data set. The iterative vector median filtering allows for fast convergence for complex computations and an output which is approximate to the mean, particularly for small data sets. In addition, a method and system for registering and matching 2.5 normal maps is provided. Registration of two maps is performed by optimally aligning their normals through 2-D warping in the image plane in conjunction with a 3-D rotation of the normals. Once aligned, the average dot-product serves as a matching metric for automatic target recognition (ATR).

Abstract: An iterative approach to vector median filtering wherein the resulting median vector need not be a member of the original data set. The iterative vector median filtering allows for fast convergence for complex computations and an output which is approximate to the mean, particularly for small data sets. In addition, a method and system for registering and matching 2.5 normal maps is provided. Registration of two maps is performed by optimally aligning their normals through 2-D warping in the image plane in conjunction with a 3-D rotation of the normals. Once aligned, the average dot-product serves as a matching metric for automatic target recognition (ATR).

Abstract: A method and apparatus is disclosed for performing blind source separation using convolutive signal decorrelation. For a first embodiment, the method accumulates a length of input signal (mixed signal) that comprises a plurality of independent signals from independent signal sources. The invention then divides the length of input signal into a plurality of T-length periods (windows) and performs a discrete Fourier transform (DFT) on the signal within each T-length period. Thereafter, estimated cross-correlation values are computed using a plurality of the averaged DFT values. A total number of K cross-correlation values are computed, where each of the K values is averaged over N of the T-length periods. Using the cross-correlation values, a gradient descent process computes the coefficients of a FIR filter that will effectively separate the source signals within the input signal. A second embodiment of the invention is directed to on-line processing of the input signal—i.e.

Abstract: A method and system for improving the accuracy and timeliness of video metadata by incorporating information related to the motion of the camera as derived from the video imagery itself. Frame-to-frame correspondences are used to accurately estimate changes in camera pose. While the method and system do not require geo-registration, geo-registration results, if available, may be considered in processing the video images and generating improved camera pose estimates.

Abstract: A method and apparatus is disclosed for performing blind source separation using convolutive signal decorrelation. For a first embodiment, the method accumulates a length of input signal (mixed signal) that includes a plurality of independent signals from independent signal sources. The invention then divides the length of input signal into a plurality of T-length periods (windows) and performs a discrete Fourier transform (DFT) on the, signal within each T-length period. Thereafter, estimated cross-correlation values are computed using a plurality of the averaged DFT values. A total number of K cross-correlation values are computed, where each of the K values is averaged over N of the T-length periods. Using the cross-correlation values, a gradient descent process computes the coefficients of a finite impulse response (FIR) filter that will effectively separate the source signals within the input signal. A second embodiment of the invention is directed to on-line processing of the input signal—i.e.

Abstract: A computer system (108) that processes mixtures of signals, such as speech and noise sources derived from multiple simultaneous microphone recordings, in order to separate them into their underlying sources. A source separation routine (124) optimizes a filter structure by minimizing cross powers of multiple output channels while enforcing geometric constraints on the filter response. The geometric constraints (209, 210, 215, 217) enforce desired responses for given locations of the underlying sources, based on the assumption that the sources are localized in space.

Abstract: An apparatus and a concomitant method for modeling local and non-local information in an image to compute an image probability distribution for the image is disclosed. In one embodiment, such an image probability distribution is determined in an object recognition system.

Abstract: A signal processing apparatus and concomitant method for learning and integrating features from multiple resolutions for detecting and/or classifying objects. The signal processing apparatus comprises a hierarchical pyramid of neural networks (HPNN) having a “fine-to-coarse” structure or a combination of the “fine-to-coarse” and the “coarse-to-fine” structures.

Abstract: A method and apparatus for training and operating a neural network using gated data. The neural network is a mixture of experts that performs “soft” partitioning of a network of experts. In a specific embodiment, the technique is used to detect malignancy by analyzing skin surface potential data. In particular, the invention uses certain patient information, such as menstrual cycle information, to “gate” the expert output data into particular populations, i.e., the network is soft partitioned into the populations. An Expectation-Maximization (EM) routine is used to train the neural network using known patient information, known measured skin potential data and correct diagnosis for the particular training data and patient information. Once trained, the neural network parameters are used in a classifier for predicting breast cancer malignancy when given the patient information and skin potentials of other patients.

Abstract: A method and apparatus that performs blind source separation using convolutive signal decorrelation. More specifically, the method accumulates a length of input signal (mixed signal) that includes a plurality of independent signals from independent signal sources. The invention then divides the length of input signal into a plurality of T-length periods (windows) and performs a discrete Fourier transform (DFT) on the signal within each T-length period. Thereafter, estimated cross-correlation values are computed using a plurality of the averaged DFT values. A total number of K cross-correlation values are computed, where each of the K values is averaged over N of the T-length periods. Using the cross-correlation values, a gradient descent process computes the coefficients of a finite impulse response (FIR) filter that will effectively separate the source signals within the input signal.

Abstract: A signal processing apparatus and concomitant method for learning and integrating features from multiple resolutions for detecting and/or classifying objects are presented. Neural networks in a pattern tree structure with tree-structured descriptions of objects in terms of simple sub-patterns, are grown and trained to detect and integrate the sub-patterns. A plurality of objective functions and their approximations are presented to train the neural networks to detect sub-patterns of features of some class of objects. Objective functions for training neural networks to detect objects whose positions in the training data are uncertain and for addressing supervised learning where there are potential errors in the training data are also presented.