Label Refinement Network for Coarse-to-Fine Semantic Segmentation
arXiv:1703.00551
Abstract
We consider the problem of semantic image segmentation using deep convolutional neural networks. We propose a novel network architecture called the label refinement network that predicts segmentation labels in a coarse-to-fine fashion at several resolutions. The segmentation labels at a coarse resolution are used together with convolutional features to obtain finer resolution segmentation labels. We define loss functions at several stages in the network to provide supervisions at different stages. Our experimental results on several standard datasets demonstrate that the proposed model provides an effective way of producing pixel-wise dense image labeling.
9 pages
References in corpus (4)
Cited by in corpus (13)
- YOLOv4: Optimal Speed and Accuracy of Object Detection
- Segmentation Transformer: Object-Contextual Representations for Semantic Segmentation
- Pyramid Stereo Matching Network
- ACFNet: Attentional Class Feature Network for Semantic Segmentation
- Position, Padding and Predictions: A Deeper Look at Position Information in CNNs
- Gated Feedback Refinement Network for Coarse-to-Fine Dense Semantic Image Labeling
- SegFix: Model-Agnostic Boundary Refinement for Segmentation
- Learning Deep Representations for Semantic Image Parsing: a Comprehensive Overview
- Semantics Meet Saliency: Exploring Domain Affinity and Models for Dual-Task Prediction
- TW-SMNet: Deep Multitask Learning of Tele-Wide Stereo Matching
- Feature Binding with Category-Dependant MixUp for Semantic Segmentation and Adversarial Robustness
- DeepDualMapper: A Gated Fusion Network for Automatic Map Extraction using Aerial Images and Trajectories
- Beyond Single Stage Encoder-Decoder Networks: Deep Decoders for Semantic Image Segmentation