# Devil is in the Edges: Learning Semantic Boundaries from Noisy Annotations

arXiv: https://arxiv.org/abs/1904.07934 · HTML: https://mlsanity.com/paper/1904.07934/

## Key facts

- Cited by 32 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 16 Apr 2019.

## Built on by (most cited first, 30 of 32)

- [BigDatasetGAN: Synthesizing ImageNet with Pixel-wise Annotations](https://mlsanity.com/paper/2201.04684/) (2022)
- [Open-World Instance Segmentation: Exploiting Pseudo Ground Truth From Learned Pairwise Affinity](https://mlsanity.com/paper/2204.06107/) (2022)
- [RINDNet: Edge Detection for Discontinuity in Reflectance, Illumination, Normal and Depth](https://mlsanity.com/paper/2108.00616/) (2021)
- [A Unified Efficient Pyramid Transformer for Semantic Segmentation](https://mlsanity.com/paper/2107.14209/) (2021)
- [Adaptive Superpixel for Active Learning in Semantic Segmentation](https://mlsanity.com/paper/2303.16817/) (2023)
- [Delving into Crispness: Guided Label Refinement for Crisp Edge Detection](https://mlsanity.com/paper/2306.15172/) (2023)
- [CP-loss: Connectivity-preserving Loss for Road Curb Detection in Autonomous Driving with Aerial Images](https://mlsanity.com/paper/2107.11920/) (2021)
- [Contour-Aware Equipotential Learning for Semantic Segmentation](https://mlsanity.com/paper/2210.00223/) (2022)
- [Robust Point Cloud Segmentation with Noisy Annotations](https://mlsanity.com/paper/2212.03242/) (2022)
- [Mobile-Seed: Joint Semantic Segmentation and Boundary Detection for Mobile Robots](https://mlsanity.com/paper/2311.12651/) (2023)
- [PNT-Edge: Towards Robust Edge Detection with Noisy Labels by Learning Pixel-level Noise Transitions](https://mlsanity.com/paper/2307.14070/) (2023)
- [Transferable Attack for Semantic Segmentation](https://mlsanity.com/paper/2307.16572/) (2023)
- [Weakly Supervised Volumetric Image Segmentation with Deformed Templates](https://mlsanity.com/paper/2106.03987/) (2021)
- [One-Step Abductive Multi-Target Learning with Diverse Noisy Samples and Its Application to Tumour Segmentation for Breast Cancer](https://mlsanity.com/paper/2110.10325/) (2021)
- [Learning to Detect Semantic Boundaries with Image-level Class Labels](https://mlsanity.com/paper/2212.07579/) (2022)
- [Empirical Study of Multi-Task Hourglass Model for Semantic Segmentation Task](https://mlsanity.com/paper/2105.13531/) (2021)
- [MSP : Refine Boundary Segmentation via Multiscale Superpixel](https://mlsanity.com/paper/2112.01746/) (2021)
- [Adjusting the Ground Truth Annotations for Connectivity-Based Learning to Delineate](https://mlsanity.com/paper/2112.02781/) (2021)
- [Accelerating the creation of instance segmentation training sets through bounding box annotation](https://mlsanity.com/paper/2205.11563/) (2022)
- [Mastering Spatial Graph Prediction of Road Networks](https://mlsanity.com/paper/2210.00828/) (2022)
- [Dunhuang murals contour generation network based on convolution and self-attention fusion](https://mlsanity.com/paper/2212.00935/) (2022)
- [Attention-based fusion of semantic boundary and non-boundary information to improve semantic segmentation](https://mlsanity.com/paper/2108.02840/) (2021)
- [Polyline Generative Navigable Space Segmentation for Autonomous Visual Navigation](https://mlsanity.com/paper/2111.00063/) (2021)
- [One-Step Abductive Multi-Target Learning with Diverse Noisy Label Samples](https://mlsanity.com/paper/2201.07933/) (2021)
- [PENet: A Joint Panoptic Edge Detection Network](https://mlsanity.com/paper/2303.08848/) (2023)
- [ECT: Fine-grained Edge Detection with Learned Cause Tokens](https://mlsanity.com/paper/2308.03092/) (2023)
- [Active Label Refinement for Semantic Segmentation of Satellite Images](https://mlsanity.com/paper/2309.06159/) (2023)
- [BAISeg: Boundary Assisted Weakly Supervised Instance Segmentation](https://mlsanity.com/paper/2406.18558/) (2024)
- [Thalia: A Global, Multi-Modal Dataset for Volcanic Activity Monitoring](https://mlsanity.com/paper/2505.17782/) (2025)
- [Single Point, Full Mask: Velocity-Guided Level Set Evolution for End-to-End Amodal Segmentation](https://mlsanity.com/paper/2508.01661/) (2025)

## Abstract (arXiv)

We tackle the problem of semantic boundary prediction, which aims to identify pixels that belong to object(class) boundaries. We notice that relevant datasets consist of a significant level of label noise, reflecting the fact that precise annotations are laborious to get and thus annotators trade-off quality with efficiency. We aim to learn sharp and precise semantic boundaries by explicitly reasoning about annotation noise during training. We propose a simple new layer and loss that can be used with existing learning-based boundary detectors. Our layer/loss enforces the detector to predict a maximum response along the normal direction at an edge, while also regularizing its direction. We further reason about true object boundaries during training using a level set formulation, which allows the network to learn from misaligned labels in an end-to-end fashion. Experiments show that we improve over the CASENet backbone network by more than 4% in terms of MF(ODS) and 18.61% in terms of AP, outperforming all current state-of-the-art methods including those that deal with alignment. Furthermore, we show that our learned network can be used to significantly improve coarse segmentation labels, lending itself as an efficient way to label new data.