×
Feb 4, 2021 · This paper proposes a novel active boundary loss for semantic segmentation. It can progressively encourage the alignment between predicted boundaries and ...
This repo is the official implementation of Active Boundary Loss for Semantic Segmentation (Paper link; arXiv:2102.02696) accepted by AAAI 2022 (Oral).
In this paper, we propose a novel active boundary loss (ABL) to progressively encourage the alignment between predicted boundaries (PDBs) and ground-truth ...
People also ask
This paper proposes a novel active boundary loss for semantic segmentation. It can progressively encourage the alignment between predicted boundaries and ...
Abstract: This paper proposes a novel active boundary loss for semantic segmentation. It can progressively encourage the alignment between predicted ...
Feb 4, 2021 · Experimental results show that training with the active boundary loss can effectively improve the boundary F-score and mean ...
This paper presents a crack-boundary refinement framework (CBRF) for meticulous segmentation of HR crack images.
Feb 10, 2021 · Abstract: This paper proposes a novel active boundary loss for semantic segmentation. It can progressively encourage the alignment between ...
Feb 4, 2021 · This paper proposes a novel active boundary loss for semantic segmentation. It can progressively encourage the alignment between predicted ...
We argue that a boundary loss can mitigate the issues related to regional losses in highly unbalanced segmentation problems.