Combinatorial Optimization for Panoptic Segmentation: A Fully Differentiable Approach
This addresses the problem of simultaneous semantic and instance segmentation for computer vision applications, representing an incremental advancement by integrating combinatorial optimization with deep learning.
The authors tackled panoptic segmentation by proposing a fully differentiable architecture combining a CNN with an asymmetric multiway cut solver, which directly maximizes a smooth surrogate of the panoptic quality metric through backpropagation. Experimental results show improvements over comparable approaches on Cityscapes and COCO datasets.
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.