# Adaptive Masked Proxies for Few-Shot Segmentation

arXiv: https://arxiv.org/abs/1902.11123 · HTML: https://mlsanity.com/paper/1902.11123/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 Feb 2019.

## Built on by (most cited first, 8 of 8)

- [Few-shot Segmentation with Optimal Transport Matching and Message Flow](https://mlsanity.com/paper/2108.08518/) (2021)
- [PFENet++: Boosting Few-shot Semantic Segmentation with the Noise-filtered Context-aware Prior Mask](https://mlsanity.com/paper/2109.13788/) (2021)
- [Few-Shot Segmentation with Global and Local Contrastive Learning](https://mlsanity.com/paper/2108.05293/) (2021)
- [CRCNet: Few-shot Segmentation with Cross-Reference and Region-Global Conditional Networks](https://mlsanity.com/paper/2208.10761/) (2022)
- [Boosting Few-shot Semantic Segmentation with Transformers](https://mlsanity.com/paper/2108.02266/) (2021)
- [MaskSplit: Self-supervised Meta-learning for Few-shot Semantic Segmentation](https://mlsanity.com/paper/2110.12207/) (2021)
- [Few Shot Semantic Segmentation: a review of methodologies, benchmarks, and open challenges](https://mlsanity.com/paper/2304.05832/) (2023)
- [Correlation Weighted Prototype-based Self-Supervised One-Shot Segmentation of Medical Images](https://mlsanity.com/paper/2408.06235/) (2024)

## Abstract (arXiv)

Deep learning has thrived by training on large-scale datasets. However, in robotics applications sample efficiency is critical. We propose a novel adaptive masked proxies method that constructs the final segmentation layer weights from few labelled samples. It utilizes multi-resolution average pooling on base embeddings masked with the label to act as a positive proxy for the new class, while fusing it with the previously learned class signatures. Our method is evaluated on PASCAL-5i dataset and outperforms the state-of-the-art in the few-shot semantic segmentation. Unlike previous methods, our approach does not require a second branch to estimate parameters or prototypes, which enables it to be used with 2-stream motion and appearance based segmentation networks. We further propose a novel setup for evaluating continual learning of object segmentation which we name incremental PASCAL (iPASCAL) where our method outperforms the baseline method. Our code is publicly available at https://github.com/MSiam/AdaptiveMaskedProxies.