Please enable JavaScript.

Coggle requires JavaScript to display documents.

Super L2O - Coggle Diagram

- - - - Deep ADMM-Net for Compressive Sensing MRI
      - One Network to Solve Them All --- Solving Linear Inverse Problems using Deep Projection Models
      - AMP-Inspired Deep Networks for Sparse Linear Inverse Problems
      - Learned D-AMP: Principled Neural Network Based Compressive Image Recovery
      - Learned Primal-dual Reconstruction
      - Theoretical Linear Convergence of Unfolded ISTA and Its Practical Weights and Thresholds
      - Tradeoffs Between Convergence Speed and Reconstruction Accuracy in Inverse Problems
      - CNN-Based Projected Gradient Descent for Consistent Image Reconstruction
      - ALISTA: Analytic Weights Are As Good As Learned Weights in LISTA
      - Unrolled Optimization with Deep Priors
      - ISTA-Net: Interpretable Optimization-Inspired Deep Network for Image Compressive Sensing
    - - Gregor & Lecun, 2010
      - Learning Efficient Structured Sparse Models
      - Efﬁcient Supervised Sparse Analysis and Synthesis Operators
      - Learning Deep $\ell_0$ Encoders
      - Maximal Sparsity with Deep Networks?
      - Understanding the Learned Iterative Soft Thresholding Algorithm with matrix factorization
    - - Learning Eﬃcient Sparse and Low Rank Models
    - - Ge Rong's recent paper
      - Reasoning layers, Le Song et al.
    - - Deep Unfolding: Model-Based Inspiration of Novel Deep Architectures
      - Learning Eﬃcient Sparse and Low Rank Models
      - Bilevel Sparse Models for Polyphonic Music Transcription
      - Supervised non-euclidean sparse NMF via bilevel optimization with applications to speech enhancement
    - - Deep Unfolding: Model-Based Inspiration of Novel Deep Architectures
    - - PDE-Net
  - - - Gradient Descent
        
        Tradeoffs Between Convergence Speed and Reconstruction Accuracy in Inverse Problems
        
        CNN-Based Projected Gradient Descent for Consistent Image Reconstruction
        
        Unrolled Optimization with Deep Priors
        
        Theoretical Interpretation of Learned Step Size in Deep-Unfolded Gradient Descent
        
        Understanding Deep Architectures with Reasoning Layer
      - ISTA
        
        Naive ISTA
        
        Gregor & Lecun, 2010
        
        D3: Deep Dual-Domain Based Fast Restoration of JPEG-Compressed Images
        
        Learning A Task-Specific Deep Architecture For Clustering
        
        Theoretical Linear Convergence of Unfolded ISTA and Its Practical Weights and Thresholds
        
        ALISTA: Analytic Weights Are As Good As Learned Weights in LISTA
        
        ISTA-Net: Interpretable Optimization-Inspired Deep Network for Image Compressive Sensing
        
        AMP
        
        Onsager-corrected deep learning for sparse linear inverse problems
        
        AMP-Inspired Deep Networks for Sparse Linear Inverse Problems
        
        Learned D-AMP: Principled Neural Network Based Compressive Image Recovery
        
        (Block)-Coordinate descent
        
        Learning Efficient Structured Sparse Models
        
        Gregor & Lecun, 2010
        
        FISTA
        
        Understanding the Learned Iterative Soft Thresholding Algorithm with matrix factorization
      - Primal-dual
        
        Learned Primal-dual Reconstruction
        
        One Network to Solve Them All --- Solving Linear Inverse Problems using Deep Projection Models
        
        Deep ADMM-Net for Compressive Sensing MRI
        
        Efﬁcient Supervised Sparse Analysis and Synthesis Operators
        
        Unrolled Optimization with Deep Priors
        
        Differentiable Linearized ADMM
        
        LSALSA: accelerated source separation via learned sparse coding
      - IHT
        
        Learning Deep $\ell_0$ Encoders
        
        Maximal Sparsity with Deep Networks?
      - Half quadratic splitting
        
        Learning Deep CNN Denoiser Prior for Image Restoration
    - - Supervised Descent Method and Its Applications to Face Alignment
    - - PDE-Net
      - Other related works
    - - Deep Unfolding of a Proximal Interior Point Method for Image Restoration
  - - - Gregor & Lecun 2010
      - ........
    - - Learning A Task-Specific Deep Architecture For Clustering
      - D3: Deep Dual-Domain Based Fast Restoration of JPEG-Compressed Images
    - - One Network to Solve Them All --- Solving Linear Inverse Problems using Deep Projection Models
      - Learned D-AMP: Principled Neural Network Based Compressive Image Recovery
      - CNN-Based Projected Gradient Descent for Consistent Image Reconstruction
      - Learning Deep CNN Denoiser Prior for Image Restoration
    - - Unrolled Optimization with Deep Priors
  - - - Gregor & Lecun, 2010
      - Learning Efficient Structured Sparse Models
      - Efﬁcient Supervised Sparse Analysis and Synthesis Operators
      - Supervised Descent Method and Its Applications to Face Alignment
      - Deep Unfolding: Model-Based Inspiration of Novel Deep Architectures
      - Onsager-corrected deep learning for sparse linear inverse problems
      - Learning Deep $\ell_0$ Encoders
      - Theoretical Linear Convergence of Unfolded ISTA and Its Practical Weights and Thresholds
      - ALISTA: Analytic Weights Are As Good As Learned Weights in LISTA
    - - Learning Eﬃcient Sparse and Low Rank Models
    - - ALISTA: Analytic Weights Are As Good As Learned Weights in LISTA
    - - Many papers image restoration
  - - - D3: Deep Dual-Domain Based Fast Restoration of JPEG-Compressed Images
      - One Network to Solve Them All --- Solving Linear Inverse Problems using Deep Projection Models
      - Learned D-AMP: Principled Neural Network Based Compressive Image Recovery
      - Tradeoffs Between Convergence Speed and Reconstruction Accuracy in Inverse Problems
      - Theoretical Linear Convergence of Unfolded ISTA and Its Practical Weights and Thresholds
      - Unrolled Optimization with Deep Priors
      - Learning Deep CNN Denoiser Prior for Image Restoration
    - - Learning A Task-Specific Deep Architecture For Clustering
    - - Deep ADMM-Net for Compressive Sensing MRI
      - Learned Primal-dual Reconstruction
      - Unsupervised Deep Basis Pursuit: Learning Reconstruction without Ground-Truth Data
    - - Deep Unfolding for Communications Systems: A Survey and Some New Directions
    - - https://slim.gatech.edu/whatisslim
      - Jianwei Ma
    - - UPR: A Model-Driven Architecture for Deep Phase Retrieval
- - - - MLP ICML 2019
      - Hierarchical RNN ICML 2017
      - Sparse Update
      - Function-based Meta Training Set ICML 2017
    - - Diverse Meta Training Set ICML 2017
      - Input feature design ICML 2017/NeurIPS 201
      - Imitation Learning / Self-improving
      - Hessian-based Regularization
    - - Gradient Estimator ICML 2019
      - Random Scaling ICML 2017
      - Convex Regularization ICML 2017
      - Attention ICML 2017
      - Curriculum Learning
      - Imitation Learning / Self-improving
      - Momentum-based Input feature design ICML 2017/NeurIPS 2019
  - - - Training Loss of Evaluation Instance
      - Testing Accuracy of Evaluation Instance
      - Distance to the Optimal
      - Privacy, Fairness, Robustness, etc.
    - - Hyperparameter Tuning
      - Network Training
      - Single Variable Optimization
        
        Lasso (both model-free and model-base)
        
        non-convex Rastrigin function (model-free, maybe Bayesian)
        
        small NN (model-free)
      - Min-max Optimization
  - - - Random Scaling and Convex Reg
      - L2O Scale
      - L2o Swarm