Model-Based Deep Learning

By Nir Shlezinger, Jay Whang, Yonina C. Eldar, A. Dimakis. Proceedings of the IEEE, 2020

512 citations recorded by the source corpus.

Signal processing, communications, and control have traditionally relied on classical statistical modeling techniques. Such model-based methods utilize mathematical formulations that represent the underlying physics, prior information, and additional domain knowledge. Simple classical models are useful but sensitive to inaccuracies and may lead to poor performance when real systems display complex or dynamic behavior. On the other hand, purely data-driven approaches that are model-agnostic are becoming increasingly popular as datasets become abundant and the power of modern deep learning pipelines increases. Deep neural networks (DNNs) use generic architectures that learn to operate from data and demonstrate excellent performance, especially for supervised problems. However, DNNs typically require massive amounts of data and immense computational resources, limiting their applicability for some scenarios. In this article, we present the leading approaches for studying and designing model-based deep learning systems. These are methods that combine principled mathematical models with data-driven systems to benefit from the advantages of both approaches. Such model-based deep learning methods exploit both partial domain knowledge, via mathematical structures designed for specific problems, and learning from limited data. Among the applications detailed in our examples for model-based deep learning are compressed sensing, digital communications, and tracking in state-space models. Our aim is to facilitate the design and study of future systems at the intersection of signal processing and machine learning that incorporate the advantages of both domains.

Subfields: Artificial Intelligence, Statistical and Nonlinear Physics

View Paper: https://www.semanticscholar.org/paper/7efe8ff7db9832d8123a29389b5bbb4903c77518

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Statistical and Nonlinear Physics

Citations in the corpus

Deep Multimodal Data Fusion (2024, background): “Moreover, DNN are commonly designed to be black boxes that do not explain how to understand and characterize the prediction result and confident intervals [159].”
Learning to Optimize: A Primer and A Benchmark (2021, background): “The work in [181] reviews model-based deep learning.”
KalmanNet: Neural Network Aided Kalman Filtering for Partially Known Dynamics (2021, unknown): “Therefore, we build upon the success of our previous work in MB deep learning for signal processing and digital communication applications [24]– [27] to propose a hybrid MB/DD online recursive filter, coined KalmanNet.”
Transformers in Medical Imaging: A Survey (2022, background): “Therefore, there has been an increasing effort in the medical imaging community to amalgamate the strengths of hand-crafted and CNNs based methods resulting in the prior information-guided CNNs models [76].”