Abstract. Weighted distance measure and discriminative training are two different approaches to enhance VQ-based solutions for speaker identification. To account for varying import...
Frameworks promote design and code reuse, at a higher level of granularity. The use of frameworks is a hard task though, because usually they lack documentation and instructions o...
We analyze the quality of condensed text on LCD displays, generated with unhinted and hinted bilevel characters, with traditional anti-aliased and with perceptually-tuned grayscale...
Abstract. This paper deals with the mapping of the parallel islandbased genetic algorithm with unidirectional ring migrations to nVidia CUDA software model. The proposed mapping is...
Manipulation of high-precision voxel volume at interactive speed is still not immediately possible by known algorithms on today's workstation hardware. We propose to decompos...