Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Ge, Jinchao, Zhang, Zeyu, Phan, Minh Hieu, Zhang, Bowen, Liu, Akide, Zhao, Yang, Zhao, Shuwen
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2408.13491
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866912521750839296
author Ge, Jinchao
Zhang, Zeyu
Phan, Minh Hieu
Zhang, Bowen
Liu, Akide
Zhao, Yang
Zhao, Shuwen
author_facet Ge, Jinchao
Zhang, Zeyu
Phan, Minh Hieu
Zhang, Bowen
Liu, Akide
Zhao, Yang
Zhao, Shuwen
contents Active learning enhances annotation efficiency by selecting the most revealing samples for labeling, thereby reducing reliance on extensive human input. Previous methods in semantic segmentation have centered on individual pixels or small areas, neglecting the rich patterns in natural images and the power of advanced pre-trained models. To address these challenges, we propose three key contributions: Firstly, we introduce Entity-Superpixel Annotation (ESA), an innovative and efficient active learning strategy which utilizes a class-agnostic mask proposal network coupled with super-pixel grouping to capture local structural cues. Additionally, our method selects a subset of entities within each image of the target domain, prioritizing superpixels with high entropy to ensure comprehensive representation. Simultaneously, it focuses on a limited number of key entities, thereby optimizing for efficiency. By utilizing an annotator-friendly design that capitalizes on the inherent structure of images, our approach significantly outperforms existing pixel-based methods, achieving superior results with minimal queries, specifically reducing click cost by 98% and enhancing performance by 1.71%. For instance, our technique requires a mere 40 clicks for annotation, a stark contrast to the 5000 clicks demanded by conventional methods.
format Preprint
id arxiv_https___arxiv_org_abs_2408_13491
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle ESA: Annotation-Efficient Active Learning for Semantic Segmentation
Ge, Jinchao
Zhang, Zeyu
Phan, Minh Hieu
Zhang, Bowen
Liu, Akide
Zhao, Yang
Zhao, Shuwen
Computer Vision and Pattern Recognition
Active learning enhances annotation efficiency by selecting the most revealing samples for labeling, thereby reducing reliance on extensive human input. Previous methods in semantic segmentation have centered on individual pixels or small areas, neglecting the rich patterns in natural images and the power of advanced pre-trained models. To address these challenges, we propose three key contributions: Firstly, we introduce Entity-Superpixel Annotation (ESA), an innovative and efficient active learning strategy which utilizes a class-agnostic mask proposal network coupled with super-pixel grouping to capture local structural cues. Additionally, our method selects a subset of entities within each image of the target domain, prioritizing superpixels with high entropy to ensure comprehensive representation. Simultaneously, it focuses on a limited number of key entities, thereby optimizing for efficiency. By utilizing an annotator-friendly design that capitalizes on the inherent structure of images, our approach significantly outperforms existing pixel-based methods, achieving superior results with minimal queries, specifically reducing click cost by 98% and enhancing performance by 1.71%. For instance, our technique requires a mere 40 clicks for annotation, a stark contrast to the 5000 clicks demanded by conventional methods.
title ESA: Annotation-Efficient Active Learning for Semantic Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2408.13491