Border-SegGCN: Improving Semantic Segmentation by Refining the Border Outline using Graph Convolutional Network

TL;DR

Border-SegGCN enhances semantic segmentation accuracy by refining object borders through a graph convolutional network applied to pre-segmented outputs, outperforming previous methods on standard datasets.

Contribution

The paper introduces a novel GCN-based border refinement module integrated with existing segmentation networks, improving boundary prediction accuracy.

Findings

Achieved 81.96% mIoU on CamVid without post-processing.

Outperformed state-of-the-art methods by 0.404% mIoU on CamVid.

Validated effectiveness on CamVid and Carla datasets.

Abstract

We present Border-SegGCN, a novel architecture to improve semantic segmentation by refining the border outline using graph convolutional networks (GCN). The semantic segmentation network such as Unet or DeepLabV3+ is used as a base network to have pre-segmented output. This output is converted into a graphical structure and fed into the GCN to improve the border pixel prediction of the pre-segmented output. We explored and studied the factors such as border thickness, number of edges for a node, and the number of features to be fed into the GCN by performing experiments. We demonstrate the effectiveness of the Border-SegGCN on the CamVid and Carla dataset, achieving a test set performance of 81.96% without any post-processing on CamVid dataset. It is higher than the reported state of the art mIoU achieved on CamVid dataset by 0.404%