文章目录
- segmentation paper
-
- 1. Anti-aliasing Semantic Reconstruction for Few-Shot Semantic Segmentation
- 2. Semantic Segmentation with Generative Models: Semi-Supervised Learning and Strong Out-of-Domain Generalization
- 3. Railroad is not a Train: Saliency as Pseudo-pixel Supervision for Weakly Supervised Semantic Segmentation
- 4. HyperSeg: Patch-wise Hypernetwork for Real-time Semantic Segmentation
- 5. Self-supervised Augmentation Consistency for Adapting Semantic Segmentation
- 6. Learning Calibrated Medical Image Segmentation via Multi-rater Agreement Modeling
- Domain Adaptive Semantic Segmentation
-
- 1.Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentation
- 2.DSP: Dual Soft-Paste for Unsupervised Domain Adaptive Semantic Segmentation
- 3.DACS: Domain Adaptation via Cross-domain Mixed Sampling
- 4.Domain Adaptive Semantic Segmentation with Self-Supervised Depth Estimation
- 5. Dual Path Learning for Domain Adaptation of Semantic Segmentation
- 6.PixMatch: Unsupervised Domain Adaptation via Pixelwise Consistency Training
- 7.SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers
segmentation paper
1. Anti-aliasing Semantic Reconstruction for Few-Shot Semantic Segmentation
- 将小样本分割重新表述为语义重建的问题
- 提出了anti-aliasing semantic reconstruction (ASR)
- 将基类特征转成基向量,张成一个语义空间,即semantic span
- 在训练期间,ASR利用语义空间,最大化基向量的正交性,最小化基类semantic aliasing ,帮助reconstruct新类
- 在推理期间,ASR利用semantic filtering 抑制语义干扰,精确激活目标物体区域
- 作者 | Binghao Liu, Yao Ding, Jianbin Jiao, Xiangyang Ji, Qixiang Ye
- 单位 | 国科大;清华
- 论文 | https://link.zhihu.com/?target=https%3A//arxiv.org/abs/2106.00184
- 代码 | https://link.zhihu.com/?target=https%3A//github.com/Bibkiller/ASR
2. Semantic Segmentation with Generative Models: Semi-Supervised Learning and Strong Out-of-Domain Generalization
- 提出一个完全生成性方法进行语义分割,基于SyleGAN2,允许半监督训练,并且展示了强大的泛化性
- 作者 | Daiqing Li, Junlin Yang, Karsten Kreis, Antonio Torralba, Sanja Fidler
- 单位 | 英伟达;多伦多大学;耶鲁大学;麻省理工学院;Vector Institute
- 论文 | https://link.zhihu.com/?target=https%3A//arxiv.org/abs/2104.05833
- 主页 | https://link.zhihu.com/?target=https%3A//nv-tlabs.github.io/semanticGAN/
3. Railroad is not a Train: Saliency as Pseudo-pixel Supervision for Weakly Supervised Semantic Segmentation
- 弱监督
- 作者 | Seungho Lee, Minhyun Lee, Jongwuk Lee, Hyunjung Shim
- 单位 | 延世大学;成均馆大学
- 论文 | https://link.zhihu.com/?target=https%3A//arxiv.org/abs/2105.08965
- 代码 | https://link.zhihu.com/?targ