Learning with Known Operators reduces Maximum Training Error Bounds

Abstract

We describe an approach for incorporating prior knowledge into machine learning algorithms. We aim at applications in physics and signal processing in which we know that certain operations must be embedded into the algorithm. Any operation that allows computation of a gradient or sub-gradient towards its inputs is suited for our framework. We derive a maximal error bound for deep nets that demonstrates that inclusion of prior knowledge results in its reduction. Furthermore, we also show experimentally that known operators reduce the number of free parameters. We apply this approach to various tasks ranging from CT image reconstruction over vessel segmentation to the derivation of previously unknown imaging algorithms. As such the concept is widely applicable for many researchers in physics, imaging, and signal processing. We assume that our analysis will support further investigation of known operators in other fields of physics, imaging, and signal processing.

Paper conditionally accepted in Nature Machine Intelligence

References in corpus (4)

• Image reconstruction by domain transform manifold learning
• Solving ill-posed inverse problems using iterative deep neural networks
• Distribution Matching Losses Can Hallucinate Features in Medical Image Translation
• Deep Denoising for Hearing Aid Applications

Cited by in corpus (25)

• Big-Data Science in Porous Materials: Materials Genomics and Machine Learning
• FISTA-Net: Learning A Fast Iterative Shrinkage Thresholding Network for Inverse Problems in Imaging
• Deep learning-based synthetic-CT generation in radiotherapy and PET: a review
• Deep variational network for rapid 4D flow MRI reconstruction
• How deep is deep enough? -- Quantifying class separability in the hidden layers of deep neural networks
• MRzero -- Fully automated discovery of MRI sequences using supervised learning
• WNet: A data-driven dual-domain denoising model for sparse-view computed tomography with a trainable reconstruction layer
• Will we ever have Conscious Machines?
• An End-To-End-Trainable Iterative Network Architecture for Accelerated Radial Multi-Coil 2D Cine MR Image Reconstruction
• Known Operator Learning and Hybrid Machine Learning in Medical Imaging -- A Review of the Past, the Present, and the Future
• Multi-modal Deep Guided Filtering for Comprehensible Medical Image Processing
• Ultra Low-Parameter Denoising: Trainable Bilateral Filter Layers in Computed Tomography
• Cephalogram Synthesis and Landmark Detection in Dental Cone-Beam CT Systems
• Artifact Reduction in 3D and 4D Cone-beam Computed Tomography Images with Deep Learning -- A Review
• Rapid Exploration of Topological Band Structures using Deep Learning
• A Scoping Review of Machine Learning Applications in Power System Protection and Disturbance Management
• Trainable Joint Bilateral Filters for Enhanced Prediction Stability in Low-dose CT
• JBFnet -- Low Dose CT Denoising by Trainable Joint Bilateral Filtering
• Spatio-temporal Multi-task Learning for Cardiac MRI Left Ventricle Quantification
• Limited Parameter Denoising for Low-dose X-ray Computed Tomography Using Deep Reinforcement Learning
• Appearance Learning for Image-based Motion Estimation in Tomography
• Learning the Update Operator for 2D/3D Image Registration
• DiffRenderGAN: Addressing Training Data Scarcity in Deep Segmentation Networks for Quantitative Nanomaterial Analysis through Differentiable Rendering and Generative Modelling
• Multi-Channel Volumetric Neural Network for Knee Cartilage Segmentation in Cone-beam CT
• Joint Super-Resolution and Rectification for Solar Cell Inspection