Karen Egiazarian

TL;DR: A novel approach to solve a problem of varying window size selection for median filtering a noisy signal based on the intersection of confidence interval rule, which enables the algorithm to be spatially adaptive in such a way that its quality is close to that which one could achieve if the smoothness of the estimated signal is known in advance.

TL;DR: In this paper, a novel efficient adaptive binary arithmetic coder is proposed which is multiplication-free and requires no look-up tables, and it is shown that in comparison with the M-coder the proposed algorithm provides comparable computational complexity, less memory footprint and bitrate savings.

TL;DR: Two approaches to adaptive JPEG-based compression of color images inside digital cameras are presented and it is demonstrated that the second approach provides more accurate estimate of degrading factor characteristics, and thus, a larger compression ratio increase compared to super-high quality (SHQ) mode used in consumer digital cameras.

TL;DR: It is shown that the proposed method of blind estimation of noise variance in a single highly textured image provides approximately two times lower estimation root mean square error than other methods.

TL;DR: This work investigates data denoising in hyperspectral terahertz pulse time-domain holography and proposes a sequential application of the block-matching algorithms oriented on work in temporal and spectral domains to improve phase image reconstruction quality.