# Core-Set Selection for Data-efficient Land Cover Segmentation

arXiv: https://arxiv.org/abs/2505.01225 · HTML: https://mlsanity.com/paper/2505.01225/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 37 references that are in the map.
- First posted on arXiv 2 May 2025.

## Builds on (most cited first, 17 of 17)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://mlsanity.com/paper/2105.15203/) (2021)
- [Confident Learning: Estimating Uncertainty in Dataset Labels](https://mlsanity.com/paper/1911.00068/) (2019)
- [The Vendi Score: A Diversity Evaluation Metric for Machine Learning](https://mlsanity.com/paper/2210.02410/) (2022)
- [Batch Active Learning at Scale](https://mlsanity.com/paper/2107.14263/) (2021)
- [Coresets and Sketches](https://mlsanity.com/paper/1601.00617/) (2016)
- [Adaptive Second Order Coresets for Data-efficient Machine Learning](https://mlsanity.com/paper/2207.13887/) (2022)
- [SEN12MS-CR-TS: A Remote Sensing Data Set for Multi-modal Multi-temporal Cloud Removal](https://mlsanity.com/paper/2201.09613/) (2022)
- [In Defense of Core-set: A Density-aware Core-set Selection for Active Learning](https://mlsanity.com/paper/2206.04838/) (2022)
- [The Principles of Data-Centric AI (DCAI)](https://mlsanity.com/paper/2211.14611/) (2022)
- [Better, Not Just More: Data-Centric Machine Learning for Earth Observation](https://mlsanity.com/paper/2312.05327/) (2023)
- [There Are No Data Like More Data- Datasets for Deep Learning in Earth Observation](https://mlsanity.com/paper/2310.19231/) (2023)
- [Dynamic Multi-Context Segmentation of Remote Sensing Images based on Convolutional Networks](https://mlsanity.com/paper/1804.04020/) (2018)
- [Data Excellence for AI: Why Should You Care](https://mlsanity.com/paper/2111.10391/) (2021)
- [Is margin all you need? An extensive empirical study of active learning on tabular data](https://mlsanity.com/paper/2210.03822/) (2022)
- [ActiveSelfHAR: Incorporating Self Training into Active Learning to Improve Cross-Subject Human Activity Recognition](https://mlsanity.com/paper/2303.15107/) (2023)

## Built on by (most cited first, 3 of 3)

- [Data-Centric AI for Tropical Agricultural Mapping: Challenges, Strategies and Scalable Solutions](https://mlsanity.com/paper/2510.16207/) (2025)
- [Towards Realistic Remote Sensing Dataset Distillation with Discriminative Prototype-guided Diffusion](https://mlsanity.com/paper/2601.15829/) (2026)
- [Fidelity-Diversity-Consistency (FDC): Data Pruning for Remote Sensing Change Detection](https://mlsanity.com/paper/2608.21754/) (2026)

## Abstract (arXiv)

The increasing accessibility of remotely sensed data and their potential to support large-scale decision-making have driven the development of deep learning models for many Earth Observation tasks. Traditionally, such models rely on large datasets. However, the common assumption that larger training datasets lead to better performance tends to overlook issues related to data redundancy, noise, and the computational cost of processing massive datasets. Effective solutions must therefore consider not only the quantity but also the quality of data. Towards this, in this paper, we introduce six basic core-set selection approaches -- that rely on imagery only, labels only, or a combination of both -- and investigate whether they can identify high-quality subsets of data capable of maintaining -- or even surpassing -- the performance achieved when using full datasets for remote sensing semantic segmentation. We benchmark such approaches against two traditional baselines on three widely used land-cover classification datasets (DFC2022, Vaihingen, and Potsdam) using two different architectures (SegFormer and U-Net), thus establishing a general baseline for future works. Our experiments show that all proposed methods consistently outperform the baselines across multiple subset sizes, with some approaches even selecting core sets that surpass training on all available data. Notably, on DFC2022, a selected subset comprising only 25% of the training data yields slightly higher SegFormer performance than training with the entire dataset. This result shows the importance and potential of data-centric learning for the remote sensing domain. The code is available at https://github.com/keillernogueira/data-centric-rs-classification/.