Multiscale Deep Equilibrium Models

TL;DR

The paper introduces multiscale deep equilibrium models (MDEQ), a new class of implicit networks that efficiently learn multi-resolution features for large-scale vision tasks, matching or surpassing state-of-the-art performance.

Contribution

MDEQ is a novel implicit network architecture that solves for equilibrium points across multiple resolutions, enabling efficient multi-task learning and high performance on vision benchmarks.

Findings

MDEQ achieves competitive results on ImageNet classification.

MDEQ performs well on semantic segmentation of high-resolution images.

MDEQ uses implicit differentiation to reduce memory usage to O(1).

Abstract

We propose a new class of implicit networks, the multiscale deep equilibrium model (MDEQ), suited to large-scale and highly hierarchical pattern recognition domains. An MDEQ directly solves for and backpropagates through the equilibrium points of multiple feature resolutions simultaneously, using implicit differentiation to avoid storing intermediate states (and thus requiring only $O(1)$ memory consumption). These simultaneously-learned multi-resolution features allow us to train a single model on a diverse set of tasks and loss functions, such as using a single MDEQ to perform both image classification and semantic segmentation. We illustrate the effectiveness of this approach on two large-scale vision tasks: ImageNet classification and semantic segmentation on high-resolution images from the Cityscapes dataset. In both settings, MDEQs are able to match or exceed the performance of recent competitive computer vision models: the first time such performance and scale have been achieved by an implicit deep learning approach. The code and pre-trained models are at https://github.com/locuslab/mdeq .