Constrained Deep Learning using Conditional Gradient and Applications in Computer Vision

Abstract

A number of results have recently demonstrated the benefits of incorporating various constraints when training deep architectures in vision and machine learning. The advantages range from guarantees for statistical generalization to better accuracy to compression. But support for general constraints within widely used libraries remains scarce and their broader deployment within many applications that can benefit from them remains under-explored. Part of the reason is that Stochastic gradient descent (SGD), the workhorse for training deep neural networks, does not natively deal with constraints with global scope very well. In this paper, we revisit a classical first order scheme from numerical optimization, Conditional Gradients (CG), that has, thus far had limited applicability in training deep models. We show via rigorous analysis how various constraints can be naturally handled by modifications of this algorithm. We provide convergence guarantees and show a suite of immediate benefits that are possible -- from training ResNets with fewer layers but better accuracy simply by substituting in our version of CG to faster training of GANs with 50% fewer epochs in image inpainting applications to provably better generalization guarantees using efficiently implementable forms of recently proposed regularizers.

References in corpus (18)

• Adam: A Method for Stochastic Optimization
• MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
• TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
• ADADELTA: An Adaptive Learning Rate Method
• Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding
• An Overview of Multi-Task Learning in Deep Neural Networks
• A Survey of Model Compression and Acceleration for Deep Neural Networks
• Anatomically Constrained Neural Networks (ACNN): Application to Cardiac Image Enhancement and Segmentation
• Convolutional neural networks with low-rank regularization
• Learning Longer Memory in Recurrent Neural Networks
• No bad local minima: Data independent training error guarantees for multilayer neural networks
• Non-convex learning via Stochastic Gradient Langevin Dynamics: a nonasymptotic analysis
• Beating the Perils of Non-Convexity: Guaranteed Training of Neural Networks using Tensor Methods
• Memory Bounded Deep Convolutional Networks
• Convergence Rate of Frank-Wolfe for Non-Convex Objectives
• Global Optimality in Tensor Factorization, Deep Learning, and Beyond
• Imposing Hard Constraints on Deep Networks: Promises and Limitations
• Low-rank passthrough neural networks

Cited by in corpus (6)

• Constrained-CNN losses for weakly supervised segmentation
• Quantized Frank-Wolfe: Faster Optimization, Lower Communication, and Projection Free
• Universal Stagewise Learning for Non-Convex Problems with Convergence on Averaged Solutions
• Model-Based Domain Generalization
• Scalable Projection-Free Optimization
• Improved Complexities for Stochastic Conditional Gradient Methods under Interpolation-like Conditions