# Data Efficiency and Transfer Robustness in Biomedical Image Segmentation: A Study of Redundancy and Forgetting with Cellpose

arXiv: https://arxiv.org/abs/2511.04803 · HTML: https://mlsanity.com/paper/2511.04803/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 12 of its 20 references that are in the map.
- First posted on arXiv 6 Nov 2025.

## Builds on (most cited first, 12 of 12)

- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [Experience Replay for Continual Learning](https://mlsanity.com/paper/1811.11682/) (2018)
- [Recent Advances of Continual Learning in Computer Vision: An Overview](https://mlsanity.com/paper/2109.11369/) (2021)
- [Dataset Quantization](https://mlsanity.com/paper/2308.10524/) (2023)
- [Diminishing Uncertainty within the Training Pool: Active Learning for Medical Image Segmentation](https://mlsanity.com/paper/2101.02323/) (2021)
- [Towards Sustainable Learning: Coresets for Data-efficient Deep Learning](https://mlsanity.com/paper/2306.01244/) (2023)
- [The Multi-modality Cell Segmentation Challenge: Towards Universal Solutions](https://mlsanity.com/paper/2308.05864/) (2023)
- [A comprehensive survey on deep active learning in medical image analysis](https://mlsanity.com/paper/2310.14230/) (2023)
- [Multi-Domain Incremental Learning for Semantic Segmentation](https://mlsanity.com/paper/2110.12205/) (2021)
- [Active learning for medical image segmentation with stochastic batches](https://mlsanity.com/paper/2301.07670/) (2023)
- [Continual Learning in Medical Imaging: A Survey and Practical Analysis](https://mlsanity.com/paper/2405.13482/) (2024)
- [Evolution-aware VAriance (EVA) Coreset Selection for Medical Image Classification](https://mlsanity.com/paper/2406.05677/) (2024)

## Abstract (arXiv)

Generalist biomedical image segmentation models such as Cellpose are increasingly applied across diverse imaging modalities and cell types. However, two critical challenges remain underexplored: (1) the extent of training data redundancy and (2) the impact of cross domain transfer on model retention. In this study, we conduct a systematic empirical analysis of these challenges using Cellpose as a case study. First, to assess data redundancy, we propose a simple dataset quantization (DQ) strategy for constructing compact yet diverse training subsets. Experiments on the Cyto dataset show that image segmentation performance saturates with only 10% of the data, revealing substantial redundancy and potential for training with minimal annotations. Latent space analysis using MAE embeddings and t-SNE confirms that DQ selected patches capture greater feature diversity than random sampling. Second, to examine catastrophic forgetting, we perform cross domain finetuning experiments and observe significant degradation in source domain performance, particularly when adapting from generalist to specialist domains. We demonstrate that selective DQ based replay reintroducing just 5-10% of the source data effectively restores source performance, while full replay can hinder target adaptation. Additionally, we find that training domain sequencing improves generalization and reduces forgetting in multi stage transfer. Our findings highlight the importance of data centric design in biomedical image segmentation and suggest that efficient training requires not only compact subsets but also retention aware learning strategies and informed domain ordering. The code is available at https://github.com/MMV-Lab/biomedseg-efficiency.