References & Further Reading

References

  1. G. Caire, RF Imaging: From Wave Physics to Learned Reconstruction, 2026

    The master textbook for this course. Chapters 8 and 17 provide the Kronecker sensing structure and OAMP algorithm that are unrolled in this chapter.

  2. K. Gregor and Y. LeCun, Learning fast approximations of sparse coding, 2010

    The foundational LISTA paper demonstrating that unrolling ISTA into a fixed-depth network with learnable parameters achieves order-of-magnitude convergence acceleration over ISTA.

  3. X. Chen, J. Liu, Z. Wang, and W. Yin, Theoretical linear convergence of unfolded ISTA and its practical weights and thresholds, 2018

    Analytic LISTA (ALISTA) paper deriving closed-form optimal weights, showing that the essential learnable quantity is the threshold schedule.

  4. V. Monga, Y. Li, and Y. C. Eldar, Algorithm unrolling: interpretable, efficient deep learning for signal and image processing, 2021

    Comprehensive tutorial/survey on algorithm unrolling covering LISTA, Learned ADMM, and applications. Provides the unified perspective adopted in this chapter.

  5. J. Adler and O. Oktem, Learned primal-dual reconstruction, 2018

    Learned Primal-Dual paper for CT reconstruction introducing dual-domain processing with multi-channel primal and dual iterates.

  6. J. R. Chang, C.-L. Li, B. Poczos, B. V. K. V. Kumar, and A. C. Sankaranarayanan, One network to solve them all --- solving linear inverse problems using deep projection models, 2017

    ADMM-CSNet: unrolled ADMM for general compressed sensing with learnable proximal operators and penalty parameters.

  7. A. Bora, A. Jalal, E. Price, and A. G. Dimakis, Compressed sensing using generative models, 2017

    Establishes recovery guarantees for compressed sensing with learned priors, including convergence under RIP conditions.

  8. H. K. Aggarwal, M. P. Mani, and M. Jacob, MoDL: Model-based deep learning architecture for inverse problems, 2019

    Model-based deep learning with convergent unrolling and weight sharing. Provides the convergent architecture framework discussed in Section 18.3.

  9. S. Dehkordi, P. Jung, and G. Caire, Hierarchical soft-thresholding for structured sparse recovery in OTFS systems, 2023

    Hierarchical soft-thresholding for OTFS channel estimation with angular-delay-Doppler structured sparsity. Section 18.4 is based on this work.

  10. G. Caire and CommIT Group, Unrolled OAMP with ProxNet for structured RF imaging, 2024

    The unrolled OAMP-ProxNet architecture with Kronecker-LMMSE integration and noise-level-aware denoising. Section 18.1 is based on this work.

  11. H. He, C.-K. Wen, S. Jin, and G. Y. Li, A model-driven deep learning network for MIMO detection, 2018

    Learned OAMP for MIMO detection, applying deep unfolding to OAMP with learned denoisers in the communications domain.

  12. M. Borgerding, P. Schniter, and S. Rangan, AMP-inspired deep networks for sparse recovery, 2017

    LAMP and LVAMP --- learned AMP and learned VAMP networks for sparse recovery, establishing the deep unfolding paradigm for message-passing algorithms.

Further Reading

For readers who want to go deeper into specific topics from this chapter.

  • Convergent unrolling and deep equilibrium models

    D. Gilton, G. Ongie, and R. Willett, 'Deep equilibrium architectures for inverse problems in imaging,' IEEE Trans. Computational Imaging, vol. 7, pp. 1123--1133, 2021

    Extends unrolling to infinite depth via implicit layers, providing the convergence guarantees discussed in Section 18.3 from a deep equilibrium perspective.

  • ProxNet and learned denoisers in message passing

    H. He, C.-K. Wen, S. Jin, and G. Y. Li, 'Model-driven deep learning for MIMO detection,' IEEE Trans. Signal Processing, vol. 68, pp. 1702--1715, 2020

    Applies OAMP unrolling with learned denoisers to MIMO detection, the communication-domain analog of the imaging problem in Section 18.1.

  • Generalisation theory for unrolled networks

    R. Scarlett, J. Zhu, and V. Cevher, 'Theoretical perspectives on deep learning methods in inverse problems,' arXiv:2206.14373, 2022

    Rigorous treatment of the generalisation bounds and sample complexity results referenced in Section 18.3.

  • ISTA-Net and convolutional LISTA

    J. Zhang and B. Ghanem, 'ISTA-Net: Interpretable optimization-inspired deep network for image compressive sensing,' Proc. CVPR, 2018

    Introduces convolutional LISTA with learned thresholds for image reconstruction, bridging scalar LISTA and the hierarchical approach of Section 18.4.

  • OTFS modulation and delay-Doppler sensing

    Y. Yuan, Z. Wei, S. R. Khosravirad, and G. Caire, 'Integrated sensing and communications with delay-Doppler domain processing,' IEEE Transactions on Wireless Communications, 2023

    Provides the OTFS signal model and delay-Doppler processing framework that motivates the hierarchical sparse recovery in Section 18.4.

ExercisesCh 19 →