Robert M. Gray

From ETHW

Robert M. Gray
Robert M. Gray
Associated organizations
Stanford University
Fields of study
Speech recognition

Biography

Robert M. Gray’s collective contributions in speech processing over the last 35 years have led to major breakthroughs in speech recognition, cellular telephony and medical imaging. He led the development of structured vector quantizers at greatly reduced complexities without sacrificing performance. Specifically, he pioneered tree-structured, pruned tree-structured, finite-state, hierarchical and entropy-coded vector quantization, as well as Lagrangian methods for quantizer optimization.

The Linde-Buzo-Gray (LBG) algorithm, which he developed with two of his students, is still the benchmark with which other design algorithms are compared. His work includes an early precursor to code excited linear predictive speech coding that is widely used to design many other types of data compression methods and Lagrangian methods, which have become the standard for rate control in video coding. He pioneered methods for combined compression and classification and is considered by many to be a leader in bringing practical and effective compression methods to medical imaging including mammography.

Dr. Gray, Lucent Technologies professor of engineering at Stanford University, Calif., has written seminal papers and books in this field, among them the well-known book, “Vector Quantization and Signal Compression” that he co-authored with Allen Gersho. An IEEE Fellow, Dr. Gray is the recipient of a number of prestigious awards as well as other honors, including a Presidential Award for Excellence in Science, Mathematics, and Engineering Mentoring. Dr. Gray holds the bachelor’s and master’s degrees from the Massachusetts Institute of Technology, Cambridge and a doctorate from the University of Southern California, Los Angeles, all in electrical engineering.

Further Reading

Robert Gray Oral History, 1991

Robert Gray Oral History, 1998