Results 71 to 80 of about 2,905 (180)

Controlling semantics of diffusion‐augmented data for unsupervised domain adaptation

IET Computer Vision, Volume 19, Issue 1, January/December 2025.
This research uses diffusion models to bridge the gap between synthetic and real images for semantic segmentation training, addressing high annotation costs. Despite introduced artefacts, employing semantic controllers improves performance by up to 5% in mIoU.
Henrietta Ridley, Roberto Alcover‐Couso, Juan C. SanMiguel   +2 more
wiley   +1 more source

Structure‐Based Uncertainty Estimation for Source‐Free Active Domain Adaptation

IET Computer Vision, Volume 19, Issue 1, January/December 2025.
We propose a novel method named structure‐based uncertainty estimation model (SUEM) for SFADA. Our method introduces a novel active sample selection strategy that combines both uncertainty and diversity sampling to identify the most informative samples.
Jihong Ouyang, Zhengjie Zhang, Qingyi Meng, Jinjin Chi   +3 more
wiley   +1 more source

Co‐teacher‐guided pseudo label supervision: A semi‐supervised learning framework for muscle and adipose tissue segmentation on chest CT scans

IET Image Processing, Volume 19, Issue 1, January/December 2025.
(1) Introduction of co‐teacher‐guided pseudo‐label supervision (CTGP): A novel SSL framework that integrates co‐training and Mean‐Teacher strategies for the segmentation of muscle and adipose tissues. (2) Development of MIAugment: A tailored augmentation technique designed specifically for medical image segmentation.
Jie Yang, Yanli Liu, Xiaoyan Chen, Tianle Chen, Qi Liu   +4 more
wiley   +1 more source

MeTa Learning-Based Optimization of Unsupervised Domain Adaptation Deep Networks

Mathematics
This paper introduces a novel unsupervised domain adaptation (UDA) method, MeTa Discriminative Class-Wise MMD (MCWMMD), which combines meta-learning with a Class-Wise Maximum Mean Discrepancy (MMD) approach to enhance domain adaptation.
Hsiau-Wen Lin, Trang-Thi Ho, Ching-Ting Tu, Hwei-Jen Lin, Chen-Hsiang Yu   +4 more
doaj   +1 more source

A Comprehensive Review of U‐Net and Its Variants: Advances and Applications in Medical Image Segmentation

IET Image Processing, Volume 19, Issue 1, January/December 2025.
This paper provides a systematic summary and reference for researchers in related fields, and we look forward to designing more efficient and stable medical image segmentation network models based on the U‐Net network. ABSTRACT Medical images often exhibit low and blurred contrast between lesions and surrounding tissues, with considerable variation in ...
Wang Jiangtao, Nur Intan Raihana Ruhaiyem, Fu Panpan   +2 more
wiley   +1 more source

Simplified Neural Unsupervised Domain Adaptation

, 2019
Unsupervised domain adaptation (UDA) is the task of modifying a statistical model trained on labeled data from a source domain to achieve better performance on data from a target domain, with access to only unlabeled data in the target domain.
Miller, Timothy A
core  

EMD‐DAAN: A Wasserstein Distance‐Based Dynamic Adversarial Domain Adaptation Network Model for Breast Ultrasound Image Classification

IET Image Processing, Volume 19, Issue 1, January/December 2025.
This introduces EMD‐DAAN, a dynamic adversarial domain adaptation network model based on the Wasserstein distance. The experimental results show that the proposed EMD‐DAAN model significantly outperforms traditional adversarial domain adaptation models such as DAAN and MK‐DAAN.
Ying Wu, Hao Huang, Bo Xu
wiley   +1 more source

Multi-level domain perturbation for source-free object detection in remote sensing images

Geo-spatial Information Science
Recent advancements in cross-domain object detection have primarily relied on unsupervised domain adaptation (UDA) techniques to bridge domain gaps in remote sensing images.
Weixing Liu, Jun Liu, Xin Su, Han Nie, Bin Luo   +4 more
doaj   +1 more source

Unsupervised Person Reidentification Using Stripe‐Driven Fusion Transformer Network

IET Software, Volume 2025, Issue 1, 2025.
In recent years, some methods utilize a transformer as the backbone to model the long‐range context dependencies, reflecting a prevailing trend in unsupervised person reidentification (Re‐ID) tasks. However, they only explore the global information through interactive learning in the framework of the transformer, which ignores the learning of the part ...
Zeyu Zang, Yang Liu, Shuang Liu, Zhong Zhang, Xinshan Zhu, Jiaxing Shen   +5 more
wiley   +1 more source

A survey on person and vehicle re‐identification

IET Computer Vision, Volume 18, Issue 8, Page 1235-1268, December 2024.
This paper summarises four popular research areas in the current field of re‐identification, focusing on the current research hotspots. These areas include the multi‐task learning domain, the generalisation learning domain, the cross‐modality domain, and the optimisation learning domain.
Zhaofa Wang, Liyang Wang, Zhiping Shi, Miaomiao Zhang, Qichuan Geng, Na Jiang   +5 more
wiley   +1 more source