模式识别与人工智能
Tuesday, Apr. 8, 2025 Home      About Journal      Editorial Board      Instructions      Ethics Statement      Contact Us                   中文
Pattern Recognition and Artificial Intelligence  2023, Vol. 36 Issue (8): 701-711    DOI: 10.16451/j.cnki.issn1003-6059.202308003
Papers and Reports Current Issue| Next Issue| Archive| Adv Search |
Instance-Level Sketch-Based Image Retrieval Based on Two Stream Multi-granularity Local Alignment Network
HAN Xuekun1,2, MIAO Duoqian1,2, ZHANG Hongyun1,2, ZHANG Qixian1,2
1. College of Electronic and Information Engineering, Tongji University, Shanghai 201804;
2. Key Laboratory of Embedded System and Service Computing, Ministry of Education, Tongji University, Shanghai 201804

Download: PDF (1625 KB)   HTML (1 KB) 
Export: BibTeX | EndNote (RIS)      
Abstract  The goal of instance-level sketch-based image retrieval is to retrieve images by sketches. There is a significant modality gap and feature misalignment issue between sketches and images. In the existing methods, the modality gap between sketches and images cannot be effectively reduced, and only information at a single granularity is captured. Thus, features cannot be aligned effectively. To address these issues, a two stream multi-granularity local alignment network(TSMLA) is proposed. A two-stream feature extractor is introduced to extract both modality-shared and modality-specific local features. These features are simultaneously utilized to calculate the distance between the sketch and the image and reduce the differences between different modalities. Moreover, a multi-granularity local alignment module is adopted to pool the distance matrix at various granularities. Local features are aligned at different scales to effectively address the problem of feature misalignment. TSMLA can fully utilize the information of sketches and real images, while effectively utilizing the connections between features of different granularities. Experiments on multiple datasets validate the effectiveness of TSMLA.
Key wordsSketch-Based Image Retrieval      Feature Extraction      Feature Fusion      Cross-Modal Retrieval     
Received: 08 June 2023     
ZTFLH: TP389.1  
Fund:National Key Research and Development Program of China(No.2022YFB3104700), National Natural Science Foundation of China(No.61976158,61976160,62076182)
Corresponding Authors: MIAO Duoqian, Ph.D., professor. His research interests include machine learning, data mining, big data analysis, granular computing, artificial intelligence and text image processing.   
About author:: HAN Xuekun, master student. His research interests include image retrieval, sketch recognition and machine learning.ZHANG Hongyun, Ph.D., associate professor. Her research interests include principal curve algorithm, granular computing and fuzzy sets.Zhang Qixian, Ph.D. candidate. His research interests include person search, computer vision and object detection.
Service
E-mail this article
Add to my bookshelf
Add to citation manager
E-mail Alert
RSS
Articles by authors
HAN Xuekun
MIAO Duoqian
ZHANG Hongyun
ZHANG Qixian
Cite this article:   
HAN Xuekun,MIAO Duoqian,ZHANG Hongyun等. Instance-Level Sketch-Based Image Retrieval Based on Two Stream Multi-granularity Local Alignment Network[J]. Pattern Recognition and Artificial Intelligence, 2023, 36(8): 701-711.
URL:  
http://manu46.magtech.com.cn/Jweb_prai/EN/10.16451/j.cnki.issn1003-6059.202308003      OR     http://manu46.magtech.com.cn/Jweb_prai/EN/Y2023/V36/I8/701
Copyright © 2010 Editorial Office of Pattern Recognition and Artificial Intelligence
Address: No.350 Shushanhu Road, Hefei, Anhui Province, P.R. China Tel: 0551-65591176 Fax:0551-65591176 Email: bjb@iim.ac.cn
Supported by Beijing Magtech  Email:support@magtech.com.cn