Sub-Microwatt Analog VLSI Support Vector Machine for Pattern Classification and Sequence Estimation

An analog system-on-chip for kernel-based pattern classification and sequence estimation is presented. State transition probabilities conditioned on input data are generated by an integrated support vector machine. Dot product based kernels and support vector coefficients are implemented in analog programmable floating gate translinear circuits, and probabilities are propagated and normalized using sub-threshold current-mode circuits. A 14-input, 24-state, and 720-support vector forward decoding kernel machine is integrated on a 3mm × 3mm chip in 0.5μm CMOS technology. Experiments with the processor trained for speaker verification and phoneme sequence estimation demonstrate real-time recognition accuracy at par with floating-point software, at sub-microwatt power.

[1]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[2]  David Haussler,et al.  Probabilistic kernel regression models , 1999, AISTATS.

[3]  Eric A. Vittoz,et al.  Low-power design: ways to approach the limits , 1994, Proceedings of IEEE International Solid-State Circuits Conference - ISSCC '94.

[4]  Carver A. Mead,et al.  A single-transistor silicon synapse , 1996 .

[5]  Gert Cauwenberghs,et al.  Forward Decoding Kernel Machines: A Hybrid HMM/SVM Approach to Sequence Recognition , 2002, SVM.

[6]  H. Loeliger,et al.  Probability propagation and decoding in analog VLSI , 1998, Proceedings. 1998 IEEE International Symposium on Information Theory (Cat. No.98CH36252).

[7]  Richard P. Lippmann,et al.  A micropower analog circuit implementation of hidden Markov model state decoding , 1997 .

[8]  A. Chandrakasan,et al.  Energy-efficient DSPs for wireless sensor networks , 2002, IEEE Signal Process. Mag..

[9]  Gert Cauwenberghs,et al.  MARGIN PROPAGATION AND FORWARD DECODING IN ANALOG VLSI , 2003 .

[10]  Hervé Bourlard,et al.  Connectionist Speech Recognition: A Hybrid Approach , 1993 .

[11]  Gert Cauwenberghs,et al.  Power dissipation limits and large margin in wireless sensors , 2003, Proceedings of the 2003 International Symposium on Circuits and Systems, 2003. ISCAS '03..

[12]  David A. Johns,et al.  BiCMOS circuits for analog Viterbi decoders , 1998 .

[13]  John Cocke,et al.  Optimal decoding of linear codes for minimizing symbol error rate (Corresp.) , 1974, IEEE Trans. Inf. Theory.