Seishi Takamura

Papers

Total Citations

H-Index

About

Seishi Takamura is a leading researcher in image and video coding, with a particular focus on evolutionary computation methods for pixel prediction. His work pioneers the use of genetic programming (GP) to automatically generate dynamic pixel predictors, moving beyond the fixed algorithms used in conventional standards like JPEG and H.264. His most cited paper, "A study on an evolutionary pixel predictor and its properties" (2009, 15 citations), lays the foundation for this approach, demonstrating how GP can evolve predictors that adapt to image content. In "Automatic pixel predictor construction using an evolutionary method" (2009, 3 citations), he further refines this concept, while "Accelerating pixel predictor evolution using edge-based class separation" (2010, 2 citations) tackles the key challenge of computational complexity by introducing edge-based classification to speed up evolution. Though citation counts are modest, Takamura’s contributions are notable for their innovative fusion of evolutionary algorithms and compression science, offering a glimpse into adaptive, self-optimizing codecs. His work is especially valuable for researchers exploring AI-driven video compression and the intersection of evolutionary computation with signal processing.