Combinatorial Optimization for Panoptic Segmentation: A Fully Differentiable Approach

TL;DR

This paper introduces a fully differentiable neural network architecture for panoptic segmentation that integrates combinatorial optimization to improve performance on large-scale datasets.

Contribution

It presents a novel end-to-end trainable framework combining deep learning with combinatorial optimization for panoptic segmentation.

Findings

Improved panoptic quality on Cityscapes dataset

Effective backpropagation through the optimization problem

Demonstrated benefits of integrating combinatorial optimization with deep learning

Abstract

We propose a fully differentiable architecture for simultaneous semantic and instance segmentation (a.k.a. panoptic segmentation) consisting of a convolutional neural network and an asymmetric multiway cut problem solver. The latter solves a combinatorial optimization problem that elegantly incorporates semantic and boundary predictions to produce a panoptic labeling. Our formulation allows to directly maximize a smooth surrogate of the panoptic quality metric by backpropagating the gradient through the optimization problem. Experimental evaluation shows improvement by backpropagating through the optimization problem w.r.t. comparable approaches on Cityscapes and COCO datasets. Overall, our approach shows the utility of using combinatorial optimization in tandem with deep learning in a challenging large scale real-world problem and showcases benefits and insights into training such an architecture.