paper-with-me

Papers

Two-Step Active Learning for Instance Segmentation with Uncertainty and Diversity Sampling

2023-09-28 · Ke Yu, Stephen Albro, Giulia Desalvo, Suraj Kothawade, Abdullah Rashwan, Sasan Tavakkol, Kayhan Batmanghelich, Xiaoqi Yin

Training high-quality instance segmentation models requires an abundance of labeled images with instance masks and classifications, which is often expensive to procure. Active learning addresses this challenge by striving for optimum performance with minimal labeling cost by selecting the most informative and representative images for labeling. Despite its potential, active learning has been less explored in instance segmentation compared to other tasks like image classification, which require less labeling. In this study, we propose a post-hoc active learning algorithm that integrates uncertainty-based sampling with diversity-based sampling. Our proposed algorithm is not only simple and easy to implement, but it also delivers superior performance on various datasets. Its practical application is demonstrated on a real-world overhead imagery dataset, where it increases the labeling efficiency fivefold.

📄 PDF Abstract BibTeX arXiv:2309.16139

Code (0)

등록된 구현이 없습니다.

Tasks

Active LearningDiversityimage-classificationImage ClassificationInstance SegmentationSegmentationSemantic Segmentation

Similar Papers 제목 키워드 기반

To Label or Not to Label: Hybrid Active Learning for Neural Machine Translation

2024-03-14 · Abdul Hameed Azeemi, Ihsan Ayyub Qazi, Agha Ali Raza

Active learning (AL) techniques reduce labeling costs for training neural machine translation (NMT) models by selecting smaller representative subsets from unlabeled data for annotation. Diversity sampling techniques sel…

Active LearningDiversityDomain AdaptationMachine Translation+3

Exploring Spatial Diversity for Region-based Active Learning

2025-07-23 · Lile Cai, Xun Xu, Lining Zhang, Chuan-Sheng Foo arxiv

State-of-the-art methods for semantic segmentation are based on deep neural networks trained on large-scale labeled datasets. Acquiring such datasets would incur large annotation costs, especially for dense pixel-level p…

Semantic SegmentationActive Learning

Active Pointly-Supervised Instance Segmentation

2022-07-23 · Chufeng Tang, Lingxi Xie, Gang Zhang, Xiaopeng Zhang 외

The requirement of expensive annotations is a major burden for training a well-performed instance segmentation model. In this paper, we present an economic active learning setting, named active pointly-supervised instanc…

Active LearningInstance SegmentationSegmentationSemantic Segmentation

Active learning for medical image segmentation with stochastic batches

2023-01-18 · Mélanie Gaillochet, Christian Desrosiers, Hervé Lombaert

The performance of learning-based algorithms improves with the amount of labelled data used for training. Yet, manually annotating data is particularly difficult for medical image segmentation tasks because of the limite…

Active LearningDiversityImage SegmentationMedical Image Segmentation+2

Predictive Accuracy-Based Active Learning for Medical Image Segmentation

2024-05-01 · Jun Shi, Shulan Ruan, Ziqi Zhu, Minfan Zhao 외

Active learning is considered a viable solution to alleviate the contradiction between the high dependency of deep learning-based segmentation methods on annotated data and the expensive pixel-level annotation cost of me…

Active LearningDiversityImage SegmentationInformativeness+3