Pith. sign in

REVIEW 3 major objections 5 minor 49 references

Matching Images and Text with Multi-modal Tensor Fusion and Re-ranking

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Image-text matching improves by learning similarity directly through rank-based tensor fusion and refining results with cross-modal re-ranking.

desk verdict A competent, efficient image-text matching pipeline that combines low-rank tensor fusion with a cross-modal re-ranking step; the gains are plausible but the re-ranking algorithm is under-specified. read the letter →

arxiv 1908.04011 v2 pith:4WOLTRQD submitted 2019-08-12 cs.CV

Tan Wang , Xing Xu , Yang Yang , Alan Hanjalic , Heng Tao Shen , Jingkuan Song This is my paper · ORCID
classification cs.CV
keywords image-textmatchingcross-modalretrievaltensorfusionk-reciprocalnearestneighborsre-rankingrankinglossFlickr30kMSCOCO
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper attempts to establish that image-text matching can be solved by directly learning a scalar similarity function through rank-based tensor fusion, rather than by mapping both modalities into a shared embedding space or by treating pairs as matching/non-matching classes. The proposed MTFN-RR combines two fusion branches, image-text and text-text, at training time, and at test time refines the initial retrieval lists by checking how well the two retrieval directions agree under a k-reciprocal-neighbour rule. On Flickr30k and MSCOCO, the authors report that MTFN alone is competitive with prior state-of-the-art methods, while MTFN-RR achieves the best reported I2T and T2I recall with a model trained in about nine hours on two GPUs. The broader point is that the expensive part of cross-modal matching may be unnecessary: global features, bilinear fusion, and a cheap post-processing re-rank can close most of the gap to attention-based fine-grained models.

What carries the argument

The central object is the rank-$R$ tensor fusion block. It takes two projected feature vectors $\tilde v=W_v v$ and $\tilde t=W_t t$, forms $R$ parallel element-wise products $(W_{\tilde v}^r\tilde v)\odot(W_{\tilde t}^r\tilde t)$, sums them, and maps the sum through a fully connected layer and sigmoid to a similarity score. The same block is reused for text-text pairs, giving a unimodal similarity that exploits the fact that each image has five captions. The re-ranking rule is the k-reciprocal-neighbour principle: a text should be retrievable from its paired image and vice versa, so the position of the query in the reverse list is used to reorder the forward list. The paper's efficiency claim follows from using global features and only two branches, with the whole model trainable in about nine hours.

What would settle it

On a benchmark where each image has one caption, or where test captions are written in a different style than training captions, rerun the T2I refinement; if the T2I R@1 does not improve over the initial list, the transfer of text-text similarity is the weak link. Alternatively, inspect the initial top-K lists for ground-truth pairs: if a large share of correct pairs are not mutual top-K neighbors in both directions, the k-reciprocal premise is violated and the re-ranking rule would demote correct matches.

Watch

Extended reading notes

Core claim

The central discovery, stated in the authors' terms, is that an explicit learned similarity function $S_{IT}(I,T)=\mathrm{sigmoid}(W_m f)$ with $f=\sum_{r=1}^R (W_{\tilde v}^r \tilde v)\odot(W_{\tilde t}^r \tilde t)$ can replace the common embedding space. A second text-text branch learns $S_{TT}(T,T')$ with the same tensor fusion. At test time, for a query image, each candidate text's own T2I list is inspected; the candidate is promoted if the query appears early in that reverse list. For a query text, the query's text-text neighborhood is found, and images are promoted if their forward I2T lists contain those neighbor captions. The paper reports that this bidirectional consistency step produces the largest gains on T2I R@1, and that the whole pipeline outperforms the listed baselines on both datasets at comparable or lower training cost.

Load-bearing premise

The load-bearing premise is that every true image-text pair is mutually retrievable: the correct caption appears near the top of its image's I2T list and the image appears near the top of that caption's T2I list, with the T2I variant additionally assuming that text-text similarity learned from training captions transfers to unseen test captions.

Editorial extensions

If this is right

  • MTFN-RR can obtain the reported top recall scores on Flickr30k and MSCOCO using global image and sentence features, without the region-level attention features used by the most accurate prior methods.
  • The cross-modal re-ranking step is training-free and the paper shows it improves multiple off-the-shelf models on I2T R@1, so retrieval pipelines can gain accuracy by adding this post-processing step.
  • The T2I task benefits most from the text-text fusion branch, because the five captions per image supply transferable semantic evidence; the reported T2I R@1 on Flickr30k rises from 46.7 with MTFN to 52.0 with MTFN-RR.
  • Training the full MTFN takes about 9 hours on two GPUs with global features, versus roughly 48 to 60 hours reported for fine-grained local-feature models, so similar accuracy is achievable at much lower compute.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A consequence the authors leave implicit is that the text-text branch's T2I gain should depend on caption redundancy; a one-caption-per-image benchmark would test whether the $S_{TT}$ transfer is the mechanism behind the T2I improvement.
  • The k-reciprocal re-ranking effectively enforces a weak transitivity condition across modalities (image-text, text-text, and image-text again); one could test whether replacing the hand-crafted position reordering with a learned consistency score preserves the gains on larger datasets.
  • Because the re-ranking is unsupervised, it could be composed with newer embedding or classifier models as a drop-in post-processor, with the prediction that per-direction recall improves wherever initial lists are already roughly reciprocal.
  • The paper's fusion-strategy ablation suggests attention contributes little once multiplicative tensor fusion is present; an untested extension is whether rank-$R$ fusion plus local region-word features would combine additively with attention or saturate.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes MTFN-RR, a two-stage framework for image-text matching. In the training stage, a Multi-modal Tensor Fusion Network learns an image-text similarity function S_IT(I,T) and a text-text similarity function S_TT(T,T') by combining rank-constrained tensor fusion with a bi-directional max-margin ranking loss, rather than learning a common embedding space. In the testing stage, a cross-modal re-ranking scheme uses k-reciprocal nearest neighbors between the I2T and T2I rankings, optionally exploiting S_TT for T2I refinement. Experiments on Flickr30k and MSCOCO report recall-at-K results and show that MTFN-RR outperforms several existing methods on many metrics while training in about 9 hours using global features; the code is released.

Significance. If the reported results are reproducible, the paper makes a useful practical contribution: it demonstrates that a directly learned similarity function with low-rank tensor fusion can be competitive with, and in several T2I metrics superior to, more complex attention-based embedding methods, and it offers a generic re-ranking scheme that improves multiple off-the-shelf methods. The strengths are the use of standard benchmarks and evaluation protocols, the inclusion of ablations for fusion strategies, rank R, and the re-ranking neighborhood K, and the public code release. The significance is constrained, however, by the under-specified T2I re-ranking procedure and by an efficiency comparison that mixes global and local feature settings.

major comments (3)
  1. [Section 3.2, Eq. (15)] The T2I re-ranking procedure is not uniquely defined by the text. The condition 'T ∈ G(T_k,K′)' depends on the parameter K′, but K′ is never reported in Section 4.1, which only gives K = 15 and K = 7. It is also not stated whether G(T,K′) includes T itself; if self is excluded, the condition can fail for all candidate images, and the paper does not specify what p(I_j) should be in that case. Finally, since R'_T2I is defined as a re-ranking of only the top-K list R_T2I(T,K), the paper does not explain how this re-ranked list is merged with the remaining tail of the full ranking, which matters because K is smaller than the number of texts needed for R@10 in I2T (for example, K = 7 on MSCOCO while R@10 is reported as 97.9). These choices are not cosmetic: the reported T2I gains of MTFN-RR over MTFN (+5.7 on Flickr30k and +2.8 on MSCOCO R@1) are produced by this step, so the 'consistent state-of-the-art' claim cannot be reproduced from the paper text alone. Please specify K′, the self-inclusion convention, the fallback for undefined p, and the merge with the retained tail.
  2. [Section 4.3, 'Analysis on Model Complexity', Fig. 5 and Table 3] The efficiency claim 'much less time complexity' is not supported by the comparison as presented. The paper contrasts MTFN trained with global features (about 9 hours) with sm-LSTM and SCAN trained with local features (about 50 and 60 hours), but this conflates the architecture with the input feature type; Fig. 5 itself shows MTFN with local features takes about 48 hours, which is close to the reported times of the comparison methods. Similarly, Table 3 compares the evaluating time of MTFN (40.2 seconds) with variants that include attention mechanisms, so the time difference is not attributable to the fusion strategy alone. To make the complexity claim credible, report training and evaluation time under matched feature types and matched modules, or explicitly decompose the time spent on feature extraction, fusion, and re-ranking.
  3. [Section 3.1, 'Text-Text Fusion', Eq. (7)] The training objective for the text-text branch is under-specified. The text says 'given two sentences in a positive pair (T_p, T_q), they have the same negative sample T_h,' but it does not define what makes a pair of sentences positive (presumably captions of the same image) or how the hard negative T_h is selected for the pair. Since S_TT is a central component of the T2I re-ranking gains, the exact supervised signal and negative sampling procedure should be stated precisely.
minor comments (5)
  1. [Abstract and Table 1] The abstract's claim that 'MTFN-RR consistently achieves the state-of-the-art matching performance' is too strong: on Flickr30k I2T, MTFN-RR (R@1 65.3, R@5 88.3, R@10 93.3) is below SCAN (I2T) (R@1 67.9, R@5 89.0, R@10 94.4). The more tempered phrasing in Section 4.2 ('achieves the state-of-the-art performance in most cases') is accurate and should be used consistently.
  2. [Tables 1, 2, and 4] All results are reported as point estimates without error bars or multiple-seed statistics. Given that several comparisons are close (e.g., MTFN vs. SCAN on Flickr30k I2T, and MTFN-RR vs. SCAN on MSCOCO T2I R@5), reporting mean and standard deviation over at least three runs would strengthen the state-of-the-art claims.
  3. [Section 3.1, Eq. (5)] The notation 'arg max_{h,p}' is ambiguous; it should be made explicit that h ranges over hard negatives that are not the positive sample. The same applies to the hard-negative definitions in the text-text loss.
  4. [Table 3] The layout of Table 3 is difficult to read: the check marks under 'Fusion Strategy' are not aligned with the individual strategy columns, making it unclear which combination each row represents. Please reformat the table with explicit row labels or symbols.
  5. [Introduction and Related Work] There are several typos and minor terminology issues, e.g., 'images and and the semantics of text' in Section 1, and 'MFTN' for 'MTFN' in Section 2. Also, the introduction describes the final layer as 'fully convolutional,' while Eq. (4) uses a fully connected layer W_m; please align the terminology.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the reported results are evaluated on held-out test splits and the MTFN/RR components are not defined in terms of the target recall scores.

full rationale

The paper's central claim is that MTFN-RR achieves state-of-the-art image-text matching performance, and this is supported by standard held-out test evaluation on Flickr30k and MSCOCO using R@K and mR metrics. The learned similarity functions SIT(I,T) and STT(T,T') are trained with ranking losses (Eqs. 5 and 7) on the training split, and the reported R@K numbers are computed from the resulting rankings on the test split; no reported metric is defined in terms of a fitted parameter or re-enters the training objective. The cross-modal re-ranking scheme (Eqs. 8-15) is a post-processing procedure applied in the testing stage, using the already-trained SIT and STT together with the k-reciprocal-neighbor assumption; it is not fitted to the test recall values, nor does it presuppose the final ranking it produces. The only self-citations in the paper ([31, 36, 37]) appear in the related-work context of cross-modal retrieval and are not load-bearing for the claimed contribution. The tensor-fusion design explicitly builds on the externally published MUTAN method [4], and the re-ranking idea builds on externally published unimodal re-ranking works [8, 28, 45]; these are cited as inspiration, not as an unverified uniqueness theorem. The T2I re-ranking in Eq. 15 is under-specified in the paper text (e.g., K' is not reported and the fallback when no T_k satisfies the condition is not stated), which is a reproducibility or correctness concern, but it is not circularity: the procedure is an additional algorithmic step, not a renaming of the fitted inputs or a prediction that reduces by construction to its own training data. Overall, the derivation chain is self-contained with respect to the claimed empirical results, and no circular step can be exhibited from the paper's equations or citations.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The method uses no invented physical or mathematical entities. It relies on standard low-rank bilinear fusion, fixed pretrained features, and a k-reciprocal-neighbor assumption for re-ranking. The main hand-set numbers are the margin, tensor rank, projection dimensions, re-ranking neighborhood size, and the unreported K'.

free parameters (6)
  • margin alpha = 0.2
    Chosen empirically for the ranking loss in Eqs. 5 and 7.
  • tensor rank R = 20
    Number of rank-1 fusion terms in Eq. 3; set empirically.
  • projection dimensions dv, dt, df = 1024
    Dimensions of projected image/text features and fusion vector; set empirically.
  • RR neighbor count K = 15 (Flickr30k), 7 (MSCOCO)
    Neighborhood size in re-ranking, tuned per dataset.
  • RR text-neighbor count K' = not reported
    Number of related texts in G(T,K') in Eq. 14 is never stated, leaving the T2I re-ranking incomplete.
  • training schedule = lr 0.0001 decayed by 2 every 10 epochs; batch size 128; 50 epochs
    Adam optimizer settings chosen empirically; standard for these benchmarks.
assumptions (4)
  • standard math Low-rank factorization captures bilinear interactions between modalities as a sum of R rank-1 outer products.
    Eq. 3 restates standard low-rank bilinear pooling from MUTAN and related VQA fusion work cited in [4, 7, 17].
  • domain assumption Global features from Faster R-CNN (ResNet) and Skip-thought GRU are sufficient for state-of-the-art image-text matching when fused with the tensor network.
    Section 4.1 fixes these features; all comparisons use them, so any claimed gain is relative to this feature set.
  • domain assumption True image-text matches are k-reciprocal nearest neighbors in both retrieval directions.
    Section 3.2 states this assumption as the basis for the re-ranking scheme.
  • domain assumption Sentences annotated to the same image share semantics, so text-text similarity learned from training captions generalizes to test captions.
    Section 3.1 introduces Text-Text Fusion and Section 3.2 uses it for T2I refinement.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Matching Images and Text with Multi-modal Tensor Fusion and Re-ranking." pith.science (2026). https://pith.science/paper/4WOLTRQD

@misc{pith2026190804011,
  author       = {Pith},
  title        = {Pith review of: Matching Images and Text with Multi-modal Tensor Fusion and Re-ranking},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4WOLTRQD}},
  note         = {Machine review of arXiv:1908.04011}
}
read the original abstract

A major challenge in matching images and text is that they have intrinsically different data distributions and feature representations. Most existing approaches are based either on embedding or classification, the first one mapping image and text instances into a common embedding space for distance measuring, and the second one regarding image-text matching as a binary classification problem. Neither of these approaches can, however, balance the matching accuracy and model complexity well. We propose a novel framework that achieves remarkable matching performance with acceptable model complexity. Specifically, in the training stage, we propose a novel Multi-modal Tensor Fusion Network (MTFN) to explicitly learn an accurate image-text similarity function with rank-based tensor fusion rather than seeking a common embedding space for each image-text instance. Then, during testing, we deploy a generic Cross-modal Re-ranking (RR) scheme for refinement without requiring additional training procedure. Extensive experiments on two datasets demonstrate that our MTFN-RR consistently achieves the state-of-the-art matching performance with much less time complexity. The implementation code is available at https://github.com/Wangt-CN/MTFN-RR-PyTorch-Code.

Figures

Figures reproduced from arXiv: 1908.04011 by the authors.

Figure 1
Figure 1. Illustration of the embedding-based (top) and [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overview architecture of our proposed framework separated by training and testing parts. 1) During training, the [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. An example of our I2T Re-ranking scheme. Given [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Visualization of the fusion vector f by classification-based method and our MTFN embedding on the part of MSCOCO test set (8000 image-text pairs) with the learned linear SVM boundary. on different fusion schemes. Following the experimental setting in [7], each combinat…
Figure 5
Figure 5. Figure 5: R@1 scores of the I2T retrieval on the MSCOCO [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Quantitative results of I2T and T2I retrieval on Flickr30k and MSCOCO datasets obtained by our MTFN and MTFN-RR [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: For the proposed RR scheme: (a) Comparison of RR [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 39 canonical work pages

  1. [1]

    [n. d.]. PyTorch Open Source Toolkit. https://github.com/pytorch/pytorch

    work page

  2. [2]

    Peter Anderson, Xiaodong He, Chris Buehler, Damien Teney, Mark Johnson, Stephen Gould, and Lei Zhang. 2017. Bottom-Up and Top-Down Attention for Image Captioning and VQA. CoRR abs/1707.07998 (2017)

    arXiv 2017

  3. [3]

    Lawrence Zitnick, and Devi Parikh

    Stanislaw Antol, Aishwarya Agrawal, Jiasen Lu, Margaret Mitchell, Dhruv Batra, C. Lawrence Zitnick, and Devi Parikh. 2015. VQA: Visual Question Answering. In 2015 IEEE International Conference on Computer Vision . 2425–2433

    work page 2015

  4. [4]

    Hedi Ben-younes, Rémi Cadène, Matthieu Cord, and Nicolas Thome. 2017. MU- TAN: Multimodal Tucker Fusion for Visual Question Answering. In IEEE Interna- tional Conference on Computer Vision . 2631–2639

    work page 2017

  5. [5]

    Kyunghyun Cho, Bart van Merrienboer, Dzmitry Bahdanau, and Yoshua Ben- gio. 2014. On the Properties of Neural Machine Translation: Encoder-Decoder Approaches. In Proceedings of SSST@EMNLP 2014, Eighth Workshop on Syntax, Semantics and Structure in Statistical Translation . 103–111

    work page 2014

  6. [6]

    Fleet, Jamie Kiros, and Sanja Fidler

    Fartash Faghri, David J. Fleet, Jamie Kiros, and Sanja Fidler. 2018. VSE++: Im- proving Visual-Semantic Embeddings with Hard Negatives. In British Machine Vision Conference 2018. 12

    work page 2018

  7. [7]

    Akira Fukui, Dong Huk Park, Daylen Yang, Anna Rohrbach, Trevor Darrell, and Marcus Rohrbach. 2016. Multimodal Compact Bilinear Pooling for Visual Question Answering and Visual Grounding. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing . 457–468

    work page 2016

  8. [8]

    Jorge García, Niki Martinel, Christian Micheloni, and Alfredo Gardel Vicente

    work page

Show all 49 references
  1. [9]

    Joty, Li Niu, and Gang Wang

    Jiuxiang Gu, Jianfei Cai, Shafiq R. Joty, Li Niu, and Gang Wang. 2017. Look, Imag- ine and Match: Improving Textual-Visual Cross-Modal Retrieval with Generative Models. CoRR abs/1711.06420 (2017)

    2017 arXiv

  2. [10]

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. 2016. Deep Residual Learning for Image Recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition. 770–778

    2016

  3. [11]

    Winston H Hsu, Lyndon S Kennedy, and Shih-Fu Chang. 2006. Video search reranking via information bottleneck principle. In Proceedings of the 14th ACM international conference on Multimedia . ACM, 35–44

    2006

  4. [12]

    Mengqiu Hu, Yang Yang, Fumin Shen, Ning Xie, Richang Hong, and Heng Tao Shen. 2019. Collective Reconstructive Embeddings for Cross-modal Hashing. IEEE Transactions on Image Processing 28, 6 (2019), 2770–2784

    2019

  5. [13]

    Yan Huang, Wei Wang, and Liang Wang. 2017. Instance-Aware Image and Sentence Matching with Selective Multimodal LSTM. In 2017 IEEE Conference on Computer Vision and Pattern Recognition . 7254–7262

    2017

  6. [14]

    Yan Huang, Qi Wu, and Liang Wang. 2017. Learning Semantic Concepts and Order for Image and Sentence Matching. CoRR abs/1712.02036 (2017)

    2017 arXiv

  7. [15]

    Allan Jabri, Armand Joulin, and Laurens van der Maaten. 2016. Revisiting Visual Question Answering Baselines. In Computer Vision - ECCV 2016 - 14th European Conference. 727–739

    2016

  8. [16]

    Karpathy and L

    A. Karpathy and L. Fei-Fei. 2015. Deep visual-semantic alignments for generating image descriptions. In Proc. IEEE Conf. Computer Vision and Pattern Recognition . 3128–3137

    2015

  9. [17]

    Jin-Hwa Kim, Kyoung Woon On, Woosang Lim, Jeonghee Kim, JungWoo Ha, and Byoung-Tak Zhang. 2016. Hadamard Product for Low-rank Bilinear Pooling. CoRR abs/1610.04325 (2016)

    2016 arXiv

  10. [18]

    Kingma and Jimmy Ba

    Diederik P. Kingma and Jimmy Ba. 2014. Adam: A Method for Stochastic Opti- mization. CoRR abs/1412.6980 (2014). arXiv:1412.6980

    2014 arXiv

  11. [19]

    Zemel, Raquel Urtasun, Antonio Torralba, and Sanja Fidler

    Ryan Kiros, Yukun Zhu, Ruslan Salakhutdinov, Richard S. Zemel, Raquel Urtasun, Antonio Torralba, and Sanja Fidler. 2015. Skip-Thought Vectors. In Advances in Neural Information Processing Systems 28: Annual Conference on Neural Informa- tion Processing Systems 2015 . 3294–3302

    2015

  12. [20]

    Shamma, Michael S

    Ranjay Krishna, Yuke Zhu, Oliver Groth, Justin Johnson, Kenji Hata, Joshua Kravitz, Stephanie Chen, Yannis Kalantidis, Li-Jia Li, David A. Shamma, Michael S. Bernstein, and Fei-Fei Li. 2016. Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotatio...

    2016 arXiv

  13. [21]

    Kuang-Huei Lee, Xi Chen, Gang Hua, Houdong Hu, and Xiaodong He. 2018. Stacked Cross Attention for Image-Text Matching. CoRR abs/1803.08024 (2018)

    2018 arXiv

  14. [22]

    Qingming Leng, Ruimin Hu, Chao Liang, Yimin Wang, and Jun Chen. 2015. Person re-identification with content and context re-ranking. Multimedia Tools Appl. 74, 17 (2015), 6989–7014

    2015

  15. [23]

    Shuang Li, Tong Xiao, Hongsheng Li, Wei Yang, and Xiaogang Wang. 2017. Identity-Aware Textual-Visual Matching with Latent Co-attention. InIEEE Inter- national Conference on Computer Vision . 1908–1917

    2017

  16. [24]

    Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Dollár, and C

    Tsung-Yi Lin, Michael Maire, Serge J. Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Dollár, and C. Lawrence Zitnick. 2014. Microsoft COCO: Common Objects in Context. In Computer Vision - ECCV 2014 - 13th European Conference . 740–755

    2014

  17. [25]

    Bakker, and Michael S

    Yu Liu, Yanming Guo, Erwin M. Bakker, and Michael S. Lew. 2017. Learning a Re- current Residual Fusion Network for Multimodal Matching. In IEEE International Conference on Computer Vision . 4127–4136

    2017

  18. [26]

    Hyeonseob Nam, Jung-Woo Ha, and Jeonghee Kim. 2017. Dual Attention Net- works for Multimodal Reasoning and Matching. In 2017 IEEE Conference on Computer Vision and Pattern Recognition . 2156–2164

    2017

  19. [27]

    Zhenxing Niu, Mo Zhou, Le Wang, Xinbo Gao, and Gang Hua. 2017. Hierarchical Multimodal LSTM for Dense Visual-Semantic Embedding. In IEEE International Conference on Computer Vision . 1899–1907

    2017

  20. [28]

    Van Gool

    Danfeng Qin, Stephan Gammeter, Lukas Bossard, Till Quack, and Luc J. Van Gool. 2011. Hello neighbor: Accurate object retrieval with k-reciprocal nearest neighbors. In The 24th IEEE Conference on Computer Vision and Pattern Recognition. 777–784

    2011

  21. [29]

    Girshick, and Jian Sun

    Shaoqing Ren, Kaiming He, Ross B. Girshick, and Jian Sun. 2015. Faster R- CNN: Towards Real-Time Object Detection with Region Proposal Networks. In Advances in Neural Information Processing Systems 28: Annual Conference on Neural Information Processing Systems 2015 . 91–99

    2015

  22. [30]

    Xiaohui Shen, Zhe Lin, Jonathan Brandt, Shai Avidan, and Ying Wu. 2012. Object retrieval and localization with spatially-constrained similarity measure and k-NN re-ranking. In 2012 IEEE Conference on Computer Vision and Pattern Recognition . 3013–3020

    2012

  23. [31]

    Jingkuan Song, Yang Yang, Yi Yang, Zi Huang, and Heng Tao Shen. 2013. Inter- media hashing for large-scale retrieval from heterogeneous data sources. In Proceedings of the ACM SIGMOD International Conference on Management of Data . 785–796

    2013

  24. [32]

    Liwei Wang, Yin Li, and Svetlana Lazebnik. 2016. Learning Deep Structure- Preserving Image-Text Embeddings. In 2016 IEEE Conference on Computer Vision and Pattern Recognition. 5005–5013

    2016

  25. [33]

    Liwei Wang, Yin Li, and Svetlana Lazebnik. 2017. Learning Two-Branch Neural Networks for Image-Text Matching Tasks. CoRR abs/1704.03470 (2017)

    2017 arXiv

  26. [34]

    Shuhui Wang, Yangyu Chen, Junbao Zhuo, Qingming Huang, and Qi Tian. 2018. Joint Global and Co-Attentive Representation Learning for Image-Sentence Re- trieval. In 2018 ACM Multimedia Conference on Multimedia Conference . ACM, 1398–1406

    2018

  27. [35]

    Courville, Ruslan Salakhutdinov, Richard S

    Kelvin Xu, Jimmy Ba, Ryan Kiros, Kyunghyun Cho, Aaron C. Courville, Ruslan Salakhutdinov, Richard S. Zemel, and Yoshua Bengio. 2015. Show, Attend and Tell: Neural Image Caption Generation with Visual Attention. In Proceedings of the 32nd International Conference on Machine Lea...

    2015

  28. [36]

    Xing Xu, Li He, Huimin Lu, Lianli Gao, and Yanli Ji. 2018. Deep Adversarial Metric Learning for Cross-Modal Retrieval. World Wide Web (2018). https: //doi.org/10.1007/s11280-018-0541-x

    2018 doi

  29. [37]

    Xing Xu, Huimin Lu, Jingkuan Song, Yang Yang, Heng Tao Shen, and Xuelong Li. 2019. Ternary Adversarial Networks With Self-Supervision for Zero-Shot Cross-Modal Retrieval. IEEE Trans Cybernetics (2019). https://doi.org/10.1109/ TCYB.2019.2928180

    2019

  30. [38]

    Linjun Yang and Alan Hanjalic. 2010. Supervised reranking for web image search. In Proceedings of the 18th ACM international conference on Multimedia . ACM, 183–192

    2010

  31. [39]

    Linjun Yang and Alan Hanjalic. 2012. Prototype-based image search reranking. IEEE transactions on multimedia 14, 3 (2012), 871–882

    2012

  32. [40]

    Mang Ye, Jun Chen, Qingming Leng, Chao Liang, Zheng Wang, and Kaimin Sun

  33. [41]

    Mang Ye, Chao Liang, Yi Yu, Zheng Wang, Qingming Leng, Chunxia Xiao, Jun Chen, and Ruimin Hu. 2016. Person Re-identification via Ranking Aggregation of Similarity Pulling and Dissimilarity Pushing. IEEE Transactions on Multimedia 18, 12 (2016), 2553–2566

    2016

  34. [42]

    In International Conference on Multimedia Modeling

    Coupled-view based ranking optimization for person re-identification. In International Conference on Multimedia Modeling . Springer, 105–117

  35. [43]

    Ying Zhang and Huchuan Lu. 2018. Deep Cross-Modal Projection Learning for Image-Text Matching. In The European Conference on Computer Vision (ECCV)

    2018

  36. [44]

    Peter Young, Alice Lai, Micah Hodosh, and Julia Hockenmaier. 2014. From image descriptions to visual denotations: New similarity metrics for semantic inference over event descriptions. TACL 2 (2014), 67–78

    2014

  37. [45]

    Zhun Zhong, Liang Zheng, Donglin Cao, and Shaozi Li. 2017. Re-ranking Per- son Re-identification with k-Reciprocal Encoding. In 2017 IEEE Conference on Computer Vision and Pattern Recognition . 3652–3661

    2017

  38. [46]

    Zhedong Zheng, Liang Zheng, Michael Garrett, Yi Yang, and Yi-Dong Shen. 2017. Dual-path convolutional image-text embedding. arXiv preprint arXiv:1711.05535 (2017)

    2017 arXiv

  39. [48]

    Bolei Zhou, Yuandong Tian, Sainbayar Sukhbaatar, Arthur Szlam, and Rob Fergus

  40. [49]

    CoRR abs/1512.02167 (2015)

    Simple Baseline for Visual Question Answering. CoRR abs/1512.02167 (2015)

    2015 arXiv

  41. [2015]

    In 2015 IEEE International Conference on Computer Vision

    Person Re-Identification Ranking Optimisation by Discriminant Context Information Analysis. In 2015 IEEE International Conference on Computer Vision . 1305–1313

    2015

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.