Information Initiative at Duke (iiD), Bass Connections-Information, Society & Culture, Biostatistics and Bioinformatics, Computer Science, Electrical and Computer Engineering (ECE), Mathematics, Pratt School of Engineering, and Statistical Science
- Gross Hall, 330 -- Ahmadieh Family Grand Hall - Map
Title: Analog to digital compression
Abstract: In this talk we consider the joint effect of sampling and lossy compression by studying the minimal distortion in recovering the analog signal from a lossy compressed version of its samples. We show that the minimal distortion in this setting is attained at some critical sampling rate which is lower than the Nyquist rate, and converges to it in the limit of infinite number of bits. That is, while the Nyquist rate is the critical sampling rate required for perfect representation without quantization constraint, the optimal sampling rate subject to this constraint is lower than Nyquist.