Deep Learning: A Signal Processing Perspective
The signal processing (SP) landscape has been enriched by recent advances in artificial intelligence (AI) and machine learning (ML), especially since 2010 or so, yielding new tools for signal estimation, classification, prediction, and manipulation. Layered signal representations, nonlinear function approximation, and nonlinear signal prediction are now feasible at very large scale in both dimensionality and data size. These are leading to significant performance gains in a variety of long standing problem domains (e.g., speech, vision), as well as providing the ability to construct new classes of nonlinear functions (e.g., fusion, nonlinear filtering). In this talk we look at the rapid evolution of signal processing tools and techniques, their strengths and weaknesses, and consider emerging frontiers. From a fundamental SP perspective, open questions include robustness, adaptivity, and performance analysis. Embedding the new techniques into emerging architectures will very likely provide new systems-level solutions for a variety of applications, taking advantage of their strengths while surmounting inherent weaknesses.