# The Devil is in Low-Level Features for Cross-Domain Few-Shot Segmentation

arXiv: https://arxiv.org/abs/2503.21150 · HTML: https://mlsanity.com/paper/2503.21150/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Builds on 37 of its 49 references that are in the map.
- First posted on arXiv 27 Mar 2025.

## Builds on (most cited first, 30 of 37)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Fully Convolutional Networks for Semantic Segmentation](https://mlsanity.com/paper/1411.4038/) (2014)
- [Prototypical Networks for Few-shot Learning](https://mlsanity.com/paper/1703.05175/) (2017)
- [Pyramid Scene Parsing Network](https://mlsanity.com/paper/1612.01105/) (2016)
- [Similarity of Neural Network Representations Revisited](https://mlsanity.com/paper/1905.00414/) (2019)
- [Sharpness-Aware Minimization for Efficiently Improving Generalization](https://mlsanity.com/paper/2010.01412/) (2020)
- [Instance Normalization: The Missing Ingredient for Fast Stylization](https://mlsanity.com/paper/1607.08022/) (2016)
- [Skin Lesion Analysis Toward Melanoma Detection 2018: A Challenge Hosted by the International Skin Imaging Collaboration (ISIC)](https://mlsanity.com/paper/1902.03368/) (2019)
- [Domain Generalization by Solving Jigsaw Puzzles](https://mlsanity.com/paper/1903.06864/) (2019)
- [PANet: Few-Shot Image Semantic Segmentation with Prototype Alignment](https://mlsanity.com/paper/1908.06391/) (2019)
- [DeepGlobe 2018: A Challenge to Parse the Earth through Satellite Images](https://mlsanity.com/paper/1805.06561/) (2018)
- [Generalizing Across Domains via Cross-Gradient Training](https://mlsanity.com/paper/1804.10745/) (2018)
- [Personalize Segment Anything Model with One Shot](https://mlsanity.com/paper/2305.03048/) (2023)
- [Adaptive Prototype Learning and Allocation for Few-Shot Segmentation](https://mlsanity.com/paper/2104.01893/) (2021)
- [FSS-1000: A 1000-Class Dataset for Few-Shot Segmentation](https://mlsanity.com/paper/1907.12347/) (2019)
- [Robust and Generalizable Visual Representation Learning via Random Convolutions](https://mlsanity.com/paper/2007.13003/) (2020)
- [Few-Shot Segmentation Without Meta-Learning: A Good Transductive Inference Is All You Need?](https://mlsanity.com/paper/2012.06166/) (2020)
- [Few-Shot Segmentation via Cycle-Consistent Transformer](https://mlsanity.com/paper/2106.02320/) (2021)
- [Learning Robust Representations by Projecting Superficial Statistics Out](https://mlsanity.com/paper/1903.06256/) (2019)
- [Simpler is Better: Few-shot Semantic Segmentation with Classifier Weight Transformer](https://mlsanity.com/paper/2108.03032/) (2021)
- [Large-scale interactive object segmentation with human annotators](https://mlsanity.com/paper/1903.10830/) (2019)
- [Learning to Learn with Variational Information Bottleneck for Domain Generalization](https://mlsanity.com/paper/2007.07645/) (2020)
- [Self-Support Few-Shot Semantic Segmentation](https://mlsanity.com/paper/2207.11549/) (2022)
- [Reliability of CKA as a Similarity Measure in Deep Learning](https://mlsanity.com/paper/2210.16156/) (2022)
- [Amplitude-Phase Recombination: Rethinking Robustness of Convolutional Neural Networks in Frequency Domain](https://mlsanity.com/paper/2108.08487/) (2021)
- [Feature-Proxy Transformer for Few-Shot Segmentation](https://mlsanity.com/paper/2210.06908/) (2022)
- [Cross-Domain Few-Shot Segmentation via Iterative Support-Query Correspondence Mining](https://mlsanity.com/paper/2401.08407/) (2024)

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

- [Unify the Views: View-Consistent Prototype Learning for Few-Shot Segmentation](https://mlsanity.com/paper/2603.05952/) (2026)
- [Training-free Cross-domain Few-shot Segmentation via Robust Semantic Representation and Matching](https://mlsanity.com/paper/2606.24297/) (2026)
- [Inter- and Intra-image Refinement for Few Shot Segmentation](https://mlsanity.com/paper/2507.05838/) (2025)
- [Cross-Domain Few-Shot Segmentation via Ordinary Differential Equations over Time Intervals](https://mlsanity.com/paper/2509.01299/) (2025)
- [Selective, Regularized, and Calibrated: Harnessing Vision Foundation Models for Cross-Domain Few-Shot Semantic Segmentation](https://mlsanity.com/paper/2605.19340/) (2026)
- [Hierarchical Spatial and Channel Aggregation for Cross-domain Few-shot Segmentation](https://mlsanity.com/paper/2606.24296/) (2026)
- [Free-Lunch Augmentation by Revisiting Diffusion-Based Data Generation for Cross-Domain Few-Shot Object Detection](https://mlsanity.com/paper/2608.04394/) (2026)
- [The First EgoCross Challenge at EgoVis 2026: Cross-Domain Egocentric Video Question Answering](https://mlsanity.com/paper/2608.04589/) (2026)

## Abstract (arXiv)

Cross-Domain Few-Shot Segmentation (CDFSS) is proposed to transfer the pixel-level segmentation capabilities learned from large-scale source-domain datasets to downstream target-domain datasets, with only a few annotated images per class. In this paper, we focus on a well-observed but unresolved phenomenon in CDFSS: for target domains, particularly those distant from the source domain, segmentation performance peaks at the very early epochs, and declines sharply as the source-domain training proceeds. We delve into this phenomenon for an interpretation: low-level features are vulnerable to domain shifts, leading to sharper loss landscapes during the source-domain training, which is the devil of CDFSS. Based on this phenomenon and interpretation, we further propose a method that includes two plug-and-play modules: one to flatten the loss landscapes for low-level features during source-domain training as a novel sharpness-aware minimization method, and the other to directly supplement target-domain information to the model during target-domain testing by low-level-based calibration. Extensive experiments on four target datasets validate our rationale and demonstrate that our method surpasses the state-of-the-art method in CDFSS signifcantly by 3.71% and 5.34% average MIoU in 1-shot and 5-shot scenarios, respectively.