A New Initial Codebook Algorithm for Learning Vector Quantization

—In this paper, a new sorting initial codebook algorithm for learning vector quantization (LVQ) based upon selforganizing feature maps (SOM) has been proposed. The basic idea is that the similar codewords are put together in the initial codebook. To improve the performance of SOM algorithm, we have presented a new winning self-organizing feature maps (WSOM). Experimental results for image VQ show that the average PSNR improvement of this new initial codebook algorithm is 0.7dB compared to the common random sampling algorithm, and the WSOM algorithm has better coding performance than the LBG and SOM algorithm. Keywords-self-organizing feature maps; learning vector quantization; LBG algorithm; image coding

• Type: Conference
• Year: 2007
• Where: ICC
• Authors: Hongsong Li, Baohua Xu
• Keywords: Communications, Feature Maps, ICC 2007, Initial Codebook Algorithm, SOM Algorithm