The discipline of signal processing: Part 2 [reflections]

The field of signal processing has evolved and diversified tremendously in its fundamentals and applications. While initiated and nurtured in electrical engineering, the discipline has been richly influenced by theories and tools of mathematics, statistics, speech and language, computer and computational sciences, physics, and geophysics. Key in the diversification of signal processing is the notion and character of a signal itself, which now is any representation of information. Li Deng, principal researcher of Microsoft Research, and affiliate professor of the University of Washington, defines that with signal processing, such rich sets of information can be transformed to enable devices or machines to operate according to the users. He expects a great deal of and growing interplay between the community of signal processing and those from artificial intelligence, machine learning, computer science, and applied mathematics. In particular, machine learning will be more and more deeply ingrained within signal processing technology.