Robust Vision Challenge 2020 -- 1st Place Report for Panoptic Segmentation

TL;DR

This paper details the winning lightweight panoptic segmentation architecture EffPS_b1bs4_RVC, which combines a shared backbone, task-specific heads, and an adaptive fusion module, achieving top results in the Robust Vision Challenge 2020.

Contribution

The paper introduces a novel, efficient panoptic segmentation architecture with a shared backbone, specialized heads, and an adaptive fusion module, setting new benchmark performance in the challenge.

Findings

Achieved 1st place on Microsoft COCO, VIPER, and WildDash datasets.

Secured 2nd place on Cityscapes and Mapillary Vistas datasets.

Demonstrated state-of-the-art performance with a lightweight model.

Abstract

In this technical report, we present key details of our winning panoptic segmentation architecture EffPS_b1bs4_RVC. Our network is a lightweight version of our state-of-the-art EfficientPS architecture that consists of our proposed shared backbone with a modified EfficientNet-B5 model as the encoder, followed by the 2-way FPN to learn semantically rich multi-scale features. It consists of two task-specific heads, a modified Mask R-CNN instance head and our novel semantic segmentation head that processes features of different scales with specialized modules for coherent feature refinement. Finally, our proposed panoptic fusion module adaptively fuses logits from each of the heads to yield the panoptic segmentation output. The Robust Vision Challenge 2020 benchmarking results show that our model is ranked #1 on Microsoft COCO, VIPER and WildDash, and is ranked #2 on Cityscapes and Mapillary Vistas, thereby achieving the overall rank #1 for the panoptic segmentation task.