Hierarchical Dense Correlation Distillation for Few-Shot Segmentation-Extended Abstract (2306.15278v1)
Abstract: Few-shot semantic segmentation (FSS) aims to form class-agnostic models segmenting unseen classes with only a handful of annotations. Previous methods limited to the semantic feature and prototype representation suffer from coarse segmentation granularity and train-set overfitting. In this work, we design Hierarchically Decoupled Matching Network (HDMNet) mining pixel-level support correlation based on the transformer architecture. The self-attention modules are used to assist in establishing hierarchical dense features, as a means to accomplish the cascade matching between query and support features. Moreover, we propose a matching module to reduce train-set overfitting and introduce correlation distillation leveraging semantic correspondence from coarse resolution to boost fine-grained segmentation. Our method performs decently in experiments. We achieve 50.0% mIoU on COCO dataset one-shot setting and 56.0% on five-shot segmentation, respectively. The code will be available on the project website. We hope our work can benefit broader industrial applications where novel classes with limited annotations are required to be decently identified.
- Few-shot segmentation without meta-learning: A good transductive inference is all you need? In CVPR.
- Region rebalance for long-tailed semantic segmentation. arXiv preprint arXiv:2204.01969, 2022.
- Generalized parametric contrastive learning. arXiv preprint arXiv:2209.12400, 2022.
- Distilling the knowledge in a neural network. arXiv preprint arXiv:1503.02531.
- Guided point contrastive learning for semi-supervised point cloud semantic segmentation. In ICCV, 2021.
- Semi-supervised semantic segmentation with directional context-aware consistency. In CVPR, 2021.
- Decouplenet: Decoupled network for domain adaptive semantic segmentation. In ECCV, 2022.
- Learning what not to segment: A new perspective on few-shot segmentation. In CVPR.
- Pfenet++: Boosting few-shot semantic segmentation with the noise-filtered context-aware prior mask. arXiv preprint arXiv:2109.13788, 2021.
- Hypercorrelation squeeze for few-shot segmentation. In ICCV.
- Feature weighting and boosting for few-shot segmentation. In ICCV.
- Neighbourhood consensus networks. NeurIPS.
- One-shot learning for semantic segmentation. arXiv preprint arXiv:1709.03410, 2017.
- Offline bilingual word vectors, orthogonal transformations and the inverted softmax. In ICLR, 2017.
- Adaptive perspective distillation for semantic segmentation. IEEE Trans. Pattern Anal. Mach. Intell.
- Learning context-aware classifier for semantic segmentation. In Proceedings of the Thirty-Seventh AAAI Conference on Artificial Intelligence, 2023.
- Generalized few-shot semantic segmentation. In CVPR.
- Prior guided feature enrichment network for few-shot segmentation. TPAMI, 2020.
- Attention is all you need. NeurIPS.
- Panet: Few-shot image semantic segmentation with prototype alignment. In ICCV.
- Canet: Class-agnostic segmentation networks with iterative refinement and attentive few-shot learning. In CVPR.
- Deep learning for medical image segmentation: Tricks, challenges and future directions. CoRR.
- Few-shot segmentation via cycle-consistent transformer. NeurIPS.
- Bohao Peng (14 papers)
- Zhuotao Tian (38 papers)
- Xiaoyang Wu (28 papers)
- Chengyao Wang (7 papers)
- Shu Liu (146 papers)
- Jingyong Su (16 papers)
- Jiaya Jia (162 papers)