The paper investigates how class imbalance and label shift—common in machine vision due to abundant defect‑free images and limited storage—affect active learning for binary semantic segmentation. It introduces a reproducible simulation protocol that independently varies these two imbalance factors using two open‑source datasets. Four active‑learning strategies (random, entropy‑based, core‑set, and galaxy sampling) are compared, showing that entropy‑based and core‑set methods generally outperform random sampling even under severe imbalance, while quantifying efficiency loss under strong label shift.
By Julien Combes (SVH, Michelin), Jean-Fran\c{c}ois Coeurjolly (SVH), Alexandre Derville (Michelin)
Pixels to Tokens introduces TEAL, a token‑space efficient active learning framework that operates at the pixel level for semantic segmentation using a frozen Vision Transformer backbone. TEAL selects diverse candidate pixels via a two‑level MaxHerding strategy on multi‑layer token descriptors and refines uncertainty on a finer decoder grid, avoiding interpolation artifacts. Across multiple datasets, TEAL outperforms prior baselines under extreme label scarcity, achieving up to +24.36 mIoU with a single labeled pixel per image and reaching 97% of fully supervised performance with only ten pixels.
By Utku Cicek, Mahip Singh, Ziyao Shang, Kimathi Kaai, C Thomas, Pedro Pablo Guerrero Vela, Alexander Wong, Sirisha Rambhatla
arXiv:2609.26631v1 Announce Type: new
Abstract: Accurate ground-based cloud classification is important for atmospheric monitoring, solar-energy forecasting, aviation weather assessment, and climate...
By Esther Bou Dagher, Viktoriya Bu-Dager, Boguslaw Zegarlinski
arXiv:2610.00851v1 Announce Type: new
Abstract: Open Set Recognition (OSR) aims to enable models to accurately classify known classes while rejecting samples from unseen classes. A key challenge in O...
By Thiru Thillai Nadarasar Bahavan, Yu Xia, Sachith Seneviratne, Saman Halgamuge
arXiv:2512.12870v2 Announce Type: replace-cross
Abstract: Active Learning (AL) is commonly used in applications where labeling data is expensive or time-consuming. In practice, however, labels are of...
By Pouya Ahadi, Blair Winograd, Camille Zaug, Karunesh Arora, Lijun Wang, Kamran Paynabar
The paper introduces a low-budget active learning approach that selects a small coreset of data points for training high-accuracy models, particularly useful when labeling is expensive, such as in medical contexts. It uses features from a pretrained self-supervised model and applies entropic optimal transport—specifically the Sinkhorn divergence—as the selection criterion, enabling dimension-free sample complexity and efficient gradient-based optimization. The method combines gradient-based candidate generation with a swap-based local search, achieving superior performance over existing heuristics on image and medical datasets.
By Rim Hajal, Mathieu Besan\c{c}on, J\'er\^ome Malick
arXiv:2510.09877v4 Announce Type: replace-cross
Abstract: Over the past couple of decades, many active learning acquisition functions have been proposed, leaving practitioners with an unclear choice...
By Kangping Hu, Stephen Mussmann
arXiv:2307.12226v3 Announce Type: replace-cross
Abstract: Machine learning models -- including prominent zero-shot models -- are often trained on datasets whose labels are only a small proportion of...
By Nicholas Roberts, Xintong Li, Dyah Adila, Sonia Cromp, Tzu-Heng Huang, Jitian Zhao, Frederic Sala
arXiv:2606. 16045v1 Announce Type: new Abstract: In the data selection problem, the objective is to choose a small, representative subset of data that can be used to efficiently train a machine learning model.
By Vincent Cohen-Addad, Sasidhar Kunapuli, Vahab Mirrokni, Mahdi Nikdan, David P. Woodruff, Samson Zhou
arXiv:2609.37481v1 Announce Type: new
Abstract: Large-scale pervasive sensing increasingly relies on high-resolution satellite imagery, yet task-specific onboard vision is constrained by costly annot...
By Ahmed Abdelnaby, Mohamed Elmahallawy, Marius Bernahrndt, Tobias Hecking
IsoAct is a post‑hoc debiasing framework that respects the geometry of non‑Euclidean representations. It identifies nuisance‑predictive directions via Fisher discriminant analysis in tangent coordinates and corrects them using manifold‑native actions that keep the representation on its manifold. Experiments on hyperspherical, hyperbolic, and SE(3) spaces show that IsoAct reduces nuisance recoverability while preserving utility and manifold validity, notably improving kinase‑group recoverability and structural prediction on the KLIFS dataset.
By Sumin Park, Taero Kim, Minhyeong Cho, Subeen Park, Jong Hwan Lee, Kyungwoo Song
arXiv:2610.19803v1 Announce Type: new
Abstract: Active domain adaptation researches focus on alleviating the annotation cost, which comprises source data and target data labelling, as well as enhanci...
By My Tra Nguyen Huynh, Viet Cuong Ta