Multi-Modal Transformer with Global-Local Alignment for Composed Query Image Retrieval

计算机科学 图像检索 嵌入 变压器 人工智能 查询扩展 编码器 模式识别(心理学) 计算机视觉 图像(数学) 情报检索 物理 量子力学 电压 操作系统
作者
Yahui Xu,Yi Bin,Jiwei Wei,Yang Yang,Guoqing Wang,Heng Tao Shen
出处
期刊:IEEE Transactions on Multimedia [Institute of Electrical and Electronics Engineers]
卷期号:25: 8346-8357 被引量:15
标识
DOI:10.1109/tmm.2023.3235495
摘要

In this paper, we study the composed query image retrieval, which aims at retrieving the target image similar to the composed query, i.e., a reference image and the desired modification text. Compared with conventional image retrieval, this task is more challenging as it not only requires precisely aligning the composed query and target image in a common embedding space, but also simultaneously extracting related information from the reference image and modification text. In order to properly extract related information from the composed query, existing methods usually embed vision-language inputs using different feature encoders, e.g., CNN for images and LSTM/BERT for text, and then employ a complicated manually-designed composition module for learning the joint image-text representation. However, the architecture discrepancy in feature encoders would restrict the vision-language plenitudinous interaction. Meanwhile, certain complicated composition designs might significantly hamper the generalization ability of the model. To tackle these problems, we propose a new framework termed ComqueryFormer, which effectively processes the composed query with the Transformer for this task. Specifically, to eliminate the architecture discrepancy, we leverage a unified transformer-based architecture to homogeneously encode the vision-language inputs. Meanwhile, instead of the complicated composition module, the neat yet effective cross-modal transformer is adopted to hierarchically fuse the composed query at various vision scales. On the other hand, we introduce an efficient global-local alignment module to narrow the distance between the composed query and the target image. It not only considers the divergence in the global joint embedding space but also forces the model to focus on the local detail differences. Extensive experiments on three real-world datasets demonstrate the superiority of our ComqueryFormer. Our code can be found at: https://github.com/uestc-xyh/ComqueryFormer .
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
PDF的下载单位、IP信息已删除 (2025-6-4)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
lh发布了新的文献求助10
1秒前
Einson完成签到 ,获得积分10
2秒前
lx发布了新的文献求助10
2秒前
001完成签到,获得积分10
3秒前
开着飞机骑拖拉机完成签到,获得积分10
3秒前
寇婧怡完成签到 ,获得积分10
4秒前
阿湫发布了新的文献求助10
4秒前
Qsss发布了新的文献求助10
4秒前
4秒前
5秒前
JamesPei应助111采纳,获得10
5秒前
执笔完成签到,获得积分10
5秒前
手可摘星辰完成签到 ,获得积分10
6秒前
量子星尘发布了新的文献求助10
6秒前
李健应助大帅采纳,获得10
7秒前
8秒前
冷艳的火龙果完成签到,获得积分10
8秒前
不知完成签到 ,获得积分10
8秒前
Zard发布了新的文献求助10
10秒前
清仔发布了新的文献求助10
10秒前
11秒前
大地上的鱼完成签到,获得积分10
11秒前
11秒前
上官若男应助平常的路人采纳,获得10
11秒前
小花发布了新的文献求助10
12秒前
庸俗完成签到,获得积分10
13秒前
14秒前
论文顺利发布了新的文献求助10
14秒前
14秒前
砚行书完成签到,获得积分10
14秒前
CodeCraft应助Qsss采纳,获得10
14秒前
情怀应助葫芦娃采纳,获得10
15秒前
小慈爱鸡完成签到 ,获得积分10
15秒前
ttelsa完成签到,获得积分10
15秒前
年轻小之完成签到 ,获得积分10
15秒前
15秒前
snowdream发布了新的文献求助10
16秒前
xiaoying完成签到,获得积分10
16秒前
18秒前
大帅发布了新的文献求助10
18秒前
高分求助中
【提示信息,请勿应助】关于scihub 10000
Les Mantodea de Guyane: Insecta, Polyneoptera [The Mantids of French Guiana] 3000
徐淮辽南地区新元古代叠层石及生物地层 3000
The Mother of All Tableaux: Order, Equivalence, and Geometry in the Large-scale Structure of Optimality Theory 3000
Global Eyelash Assessment scale (GEA) 1000
Picture Books with Same-sex Parented Families: Unintentional Censorship 550
Research on Disturbance Rejection Control Algorithm for Aerial Operation Robots 500
热门求助领域 (近24小时)
化学 材料科学 医学 生物 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 遗传学 基因 物理化学 催化作用 冶金 细胞生物学 免疫学
热门帖子
关注 科研通微信公众号,转发送积分 4038524
求助须知:如何正确求助?哪些是违规求助? 3576221
关于积分的说明 11374737
捐赠科研通 3305912
什么是DOI,文献DOI怎么找? 1819354
邀请新用户注册赠送积分活动 892688
科研通“疑难数据库(出版商)”最低求助积分说明 815048