A Real-Time Learning Processor Based on K-means Algorithm with Automatic Seeds Generation

A full-custom learning processor architecture has been developed based on the K-means algorithm aiming at realtime clustering applications. In order to accelerate the convergence and improve the quality of solutions, an automatic initial seeds generation function has been implemented in the architecture. The concept has been verified by the measurement of the proof-of-concept chip designed and fabricated in a 0.18-mum 5-metal CMOS technology. A full custom chip was also designed using the same technology and sent to fabrication and its operation was confirmed by simulation.

[1]  Orly Yadid-Pecht,et al.  Hardware-driven adaptive k-means clustering for real-time video imaging , 2005 .

[2]  Maya Gokhale,et al.  Applying reconfigurable hardware to the analysis of multispectral and hyperspectral imagery , 2002, SPIE Optics + Photonics.

[3]  T. Shibata,et al.  A general-purpose vector-quantization processor employing two-dimensional bit-propagating winner-take-all , 2002, 2002 Symposium on VLSI Circuits. Digest of Technical Papers (Cat. No.02CH37302).

[4]  C.-C. Jay Kuo,et al.  A new initialization technique for generalized Lloyd iteration , 1994, IEEE Signal Processing Letters.

[5]  Man Lan,et al.  Initialization of cluster refinement algorithms: a review and comparative study , 2004, 2004 IEEE International Joint Conference on Neural Networks (IEEE Cat. No.04CH37541).

[6]  J. MacQueen.  Some methods for classification and analysis of multivariate observations , 1967 .

[7]  T. Morimoto,et al.  A fully-parallel vector quantization processor for real-time motion picture compression , 1997, 1997 IEEE International Solids-State Circuits Conference. Digest of Technical Papers.

[8]  Abel G. Silva-Filho,et al.  Hyperspectral images clustering on reconfigurable hardware using the k-means algorithm , 2003, 16th Symposium on Integrated Circuits and Systems Design, 2003. SBCCI 2003. Proceedings..

[9]  Dmitriy Fradkin,et al.  Image compression in real-time multiprocessor systems using divisive K-means clustering , 2003, IEMC '03 Proceedings. Managing Technologically Driven Organizations: The Human Side of Innovation and Change (IEEE Cat. No.03CH37502).

[10]  Tadashi Shibata,et al.  An image representation algorithm compatible with neural-associative-processor-based hardware recognition systems , 2003, IEEE Trans. Neural Networks.