Only Classification Head Is Sufficient for Medical Image Segmentation DOI

Hongbin Wei,

Zhiwei Hu, Bo Chen

et al.

Lecture notes in computer science, Journal Year: 2023, Volume and Issue: unknown, P. 296 - 308

Published: Dec. 25, 2023

Language: Английский

Stealth sight: A multi perspective approach for camouflaged object detection DOI
S. Domnic,

J. S.

Image and Vision Computing, Journal Year: 2025, Volume and Issue: unknown, P. 105517 - 105517

Published: March 1, 2025

Language: Английский

Citations

0
G2LNet: Global to local information communication for camouflaged object detection DOI
Na Wang, Guanhua Zhang, Wendong Wang

et al.

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 127507 - 127507

Published: April 1, 2025

Language: Английский

Citations

0
Isomer: Isomerous Transformer for Zero-shot Video Object Segmentation DOI
Yichen Yuan, Yifan Wang, Lijun Wang

et al.

2021 IEEE/CVF International Conference on Computer Vision (ICCV), Journal Year: 2023, Volume and Issue: unknown, P. 966 - 976

Published: Oct. 1, 2023

Recent leading zero-shot video object segmentation (ZVOS) works devote to integrating appearance and motion information by elaborately designing feature fusion modules identically applying them in multiple stages. Our preliminary experiments show that with the strong long-range dependency modeling capacity of Transformer, simply concatenating two modality features feeding vanilla Transformers for can distinctly benefit performance but at a cost heavy computation. Through further empirical analysis, we find attention dependencies learned Transformer different stages exhibit completely properties: global query-independent low-level semantic-specific high-level Motivated observations, propose variants: i) Context-Sharing (CST) learns global-shared contextual within image frames lightweight ii) Semantic Gathering-Scattering (SGST) models semantic correlation separately foreground background reduces computation soft token merging mechanism. We apply CST SGST fusions, respectively, formulating level-isomerous framework ZVOS task. Compared baseline uses multi-stage fusion, ours significantly increase speed 13× achieves new state-of-the-art performance. Code is available https://github.com/DLUT-yyc/Isomer.

Language: Английский

Citations

8
Open-Vocabulary Camouflaged Object Segmentation DOI
Youwei Pang, Xiaoqi Zhao,

Jiaming Zuo

et al.

Lecture notes in computer science, Journal Year: 2024, Volume and Issue: unknown, P. 476 - 495

Published: Nov. 22, 2024

Language: Английский

Citations

1
Growth Simulation Network for Polyp Segmentation DOI

Hongbin Wei,

Xiaoqi Zhao,

Long Lv

et al.

Lecture notes in computer science, Journal Year: 2023, Volume and Issue: unknown, P. 3 - 15

Published: Dec. 25, 2023

Language: Английский

Citations

1
Towards Automatic Power Battery Detection: New Challenge, Benchmark Dataset and Baseline DOI
Xiaoqi Zhao, Youwei Pang, Zhenyu Chen

et al.

2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Journal Year: 2024, Volume and Issue: 11, P. 22020 - 22029

Published: June 16, 2024

Language: Английский

Citations

0
Advancing a Vision Foundation Model for Ming-Style Furniture Image Segmentation: A New Dataset and Method DOI Creative Commons
Yuehua Wan,

W. Wang,

Meng Zhang

et al.

Sensors, Journal Year: 2024, Volume and Issue: 25(1), P. 96 - 96

Published: Dec. 27, 2024

This paper tackles the challenge of accurately segmenting images Ming-style furniture, an important aspect China’s cultural heritage, to aid in its preservation and analysis. Existing vision foundation models, like segment anything model (SAM), struggle with complex structures Ming furniture due need for manual prompts imprecise segmentation outputs. To address these limitations, we introduce two key innovations: material attribute prompter (MAP), which automatically generates based on furniture’s properties, structure refinement module (SRM), enhances by combining high- low-level features. Additionally, present MF2K dataset, includes 2073 annotated pixel-level masks across eight materials environments. Our experiments demonstrate that proposed method significantly improves accuracy, outperforming state-of-the-art models terms mean intersection over union (mIoU). Ablation studies highlight contributions MAP SRM both performance computational efficiency. work offers a powerful automated solution intricate structures, facilitating digital in-depth analysis furniture.

Language: Английский

Citations

0
Only Classification Head Is Sufficient for Medical Image Segmentation DOI

Hongbin Wei,

Zhiwei Hu, Bo Chen

et al.

Lecture notes in computer science, Journal Year: 2023, Volume and Issue: unknown, P. 296 - 308

Published: Dec. 25, 2023

Language: Английский

Citations

0