Recognition using information-optimal adaptive feature-specific imaging.

We present an information-theoretic adaptive feature-specific imaging (AFSI) system for a M-class recognition task. The proposed system utilizes the recently developed task-specific information (TSI) framework to incorporate the knowledge from previous measurements and adapt the projection matrix at each step. The decision-making framework is based on sequential hypothesis testing. We quantify the number of measurements required to achieve a specified probability of misclassification (P(e)), and we compare the performances of three approaches: the new TSI-based AFSI system, a previously reported statistical AFSI system, and static FSI (SFSI). The TSI-based AFSI system exhibits significant improvement compared with SFSI and statistical AFSI at low signal-to-noise ratio (SNR). It is shown that for M=4 hypotheses, SNR=-20 dB and desired P(e)=10(-2), TSI-based AFSI requires 3 times fewer measurements than statistical AFSI, and 16 times fewer measurements than SFSI. We also describe an extension of the proposed method that is suitable for recognition in the presence of nuisance parameters such as illumination conditions and target orientations.

[1]  Alan L. Yuille,et al.  Feature extraction from faces using deformable templates , 2004, International Journal of Computer Vision.

[2]  Richard G. Baraniuk,et al.  A new compressive imaging camera architecture using optical-domain compression , 2006, Electronic Imaging.

[3]  B. V. Vijaya Kumar,et al.  Unconstrained correlation filters. , 1994, Applied optics.

[4]  Matti Pietikäinen,et al.  Physics-based face database for color research , 2000, J. Electronic Imaging.

[5]  Xiaobai Sun,et al.  The Quantized Cosine Transform for Sensor-Layer Image Compression , 2005 .

[6]  Mark A Neifeld,et al.  Feature-specific structured imaging. , 2006, Applied optics.

[7]  Jannick P. Rolland,et al.  Linear discriminants and image quality , 1991, Image Vis. Comput..

[8]  David J. Kriegman,et al.  Eigenfaces vs. Fisherfaces: Recognition Using Class Specific Linear Projection , 1996, ECCV.

[9]  Azriel Rosenfeld,et al.  Face recognition: A literature survey , 2003, CSUR.

[10]  P. Armitage.  Sequential Analysis with More than Two Alternative Hypotheses, and its Relation to Discriminant Function Analysis , 1950 .

[11]  Ah Chung Tsoi,et al.  Face recognition: a convolutional neural-network approach , 1997, IEEE Trans. Neural Networks.

[12]  Mark A Neifeld,et al.  Face recognition by using feature-specific imaging. , 2005, Applied optics.

[13]  Amit Ashok,et al.  Compressive imaging system design using task-specific information. , 2008, Applied optics.

[14]  Pavel Pudil,et al.  Conditional Mutual Information Based Feature Selection for Classification Task , 2007, CIARP.

[15]  Richard G. Baraniuk,et al.  Sparse Signal Detection from Incoherent Projections , 2006, 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings.

[16]  Jun Ke,et al.  Optical architectures for compressive imaging , 2007 .

[17]  B Javidi,et al.  Optimum receiver design for pattern recognition with nonoverlapping target and scene noise. , 1993, Optics letters.

[18]  Edward R. Dowski,et al.  A New Paradigm for Imaging Systems , 2002, PICS.

[19]  Amit Ashok,et al.  Task-specific information for imaging system analysis. , 2007, Journal of the Optical Society of America. A, Optics, image science, and vision.

[20]  A. Wald.  Sequential Tests of Statistical Hypotheses , 1945 .

[21]  Mark A Neifeld,et al.  Feature-specific imaging. , 2003, Applied optics.

[22]  Kannan Ramchandran,et al.  Information-Theoretic Bounds on Target Recognition Performance Based on Degraded Image Data , 2002, IEEE Trans. Pattern Anal. Mach. Intell..

[23]  A. B. Vander Lugt,et al.  Signal detection by complex spatial filtering , 1964, IEEE Trans. Inf. Theory.

[24]  Mark A Neifeld,et al.  Adaptive feature-specific imaging: a face recognition example. , 2008, Applied optics.

[25]  Witold Pedrycz,et al.  Face recognition: A study in information fusion using fuzzy integral , 2005, Pattern Recognit. Lett..

[26]  Mark R. Bell.  Information theory and radar waveform design , 1993, IEEE Trans. Inf. Theory.

[27]  D. Brady.  Multiplex sensors and the constant radiance theorem. , 2002, Optics letters.

[28]  Roberto Battiti,et al.  Using mutual information for selecting features in supervised neural net learning , 1994, IEEE Trans. Neural Networks.

[29]  Daniel Pérez Palomar,et al.  Gradient of mutual information in linear vector Gaussian channels , 2006, IEEE Transactions on Information Theory.

[30]  M.A. Neifeld,et al.  Adaptive Waveform Design and Sequential Hypothesis Testing for Target Recognition With Active Sensors , 2007, IEEE Journal of Selected Topics in Signal Processing.

[31]  Alex Pentland,et al.  Face recognition using eigenfaces , 1991, Proceedings. 1991 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[32]  Sudhakar Prasad,et al.  Engineering the pupil phase to improve image quality , 2003, SPIE Defense + Commercial Sensing.

[33]  S. U. Pillai,et al.  Adaptive transmission radar: the next "wave"? , 2000, Proceedings of the IEEE 2000 National Aerospace and Electronics Conference. NAECON 2000. Engineering Tomorrow (Cat. No.00CH37093).