Skip to main content
SpringerLink
Log in

Enhancing Intra-modal Similarity in a Cross-Modal Triplet Loss

  • Conference paper
  • First Online:
Discovery Science (DS 2023)

Part of the book series: Lecture Notes in Computer Science ((LNAI,volume 14276))

Included in the following conference series:

  • 558 Accesses

Abstract

Cross-modal retrieval requires building a common latent space that captures and correlates information from different data modalities, usually images and texts. Cross-modal training based on the triplet loss with hard negative mining is a state-of-the-art technique to address this problem. This paper shows that such approach is not always effective in handling intra-modal similarities. Specifically, we found that this method can lead to inconsistent similarity orderings in the latent space, where intra-modal pairs with unknown ground-truth similarity are ranked higher than cross-modal pairs representing the same concept. To address this problem, we propose two novel loss functions that leverage intra-modal similarity constraints available in a training triplet but not used by the original formulation. Additionally, this paper explores the application of this framework to unsupervised image retrieval problems, where cross-modal training can provide the supervisory signals that are otherwise missing in the absence of category labels. Up to our knowledge, we are the first to evaluate cross-modal training for intra-modal retrieval without labels.

We present comprehensive experiments on MS-COCO and Flickr30K, demonstrating the advantages and limitations of the proposed methods in cross-modal and intra-modal retrieval tasks in terms of performance and novelty measures. Our code is publicly available on GitHub https://github.com/MariodotR/FullHN.git.

This research was partially funded by National Agency for Research and Development (ANID, Chile), grant numbers FONDEF IT21I0019, ANID PIA/APOYO AFB180002 and ANID-Basal Project FB0008.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Log in via an institution
Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 69.99
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 89.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Notes

  1. 1.

    For MS-COCO, we report results for 5k images.

  2. 2.

    With results of \([ 351.3 - 350.3 , 351.0-349.3, 350.4-351.4]\), respectively.

  3. 3.

    With results of [367.0, 368.1, 375.9, 377.5, 370.4], respectively.

  4. 4.

    The best value is underlined and the best without considering TERAN is highlighted in bold.

References

  1. Chaudhuri, U., Banerjee, B., Bhattacharya, A., Datcu, M.: CMIR-NET: a deep learning based model for cross-modal retrieval in remote sensing. Pattern Recogn. Lett. 131, 456–462 (2020)

    Article  Google Scholar 

  2. Clarke, C.L., et al.: Novelty and diversity in information retrieval evaluation. In: SIGIR 2008 ,p p. 659–666. ACM, New York (2008)

    Google Scholar 

  3. Do, T.T., Tran, T., Ian, R., et al.: A theoretically sound upper bound on the triplet loss for improving the efficiency of deep distance metric learning. In: IEEE CVPR, pp. 10404–10413 (2019)

    Google Scholar 

  4. Dubey, S.R.: A decade survey of content based image retrieval using deep learning. IEEE Trans. Circ. Syst. Video Technol. 32, 2687–2704 (2020)

    Article  Google Scholar 

  5. Faghri, F., Fleet, D.J., Kiros, J.R., Fidler, S.: VSE++: improving visual-semantic embeddings with hard negatives. In: Proceedings of BMVC (2017)

    Google Scholar 

  6. Ge, W., Huang, W., Dong, D., Scott, M.R.: Deep metric learning with hierarchical triplet loss. In: Ferrari, V., Hebert, M., Sminchisescu, C., Weiss, Y. (eds.) ECCV 2018. LNCS, vol. 11210, pp. 272–288. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-01231-1_17

    Chapter  Google Scholar 

  7. Gong, Y., Cosma, G.: Improving visual-semantic embeddings by learning semantically-enhanced hard negatives for cross-modal information retrieval. Pattern Recogn. 137, 109272 (2023)

    Article  Google Scholar 

  8. Karpathy, A., Fei-Fei, L.: Deep visual-semantic alignments for generating image descriptions. IEEE Trans. Pattern Anal. Mach. Intell. 39(4), 664–676 (2017)

    Article  Google Scholar 

  9. Mahmut, K., Şakir, H.: Deep metric learning: a survey. Symmetry 11(9), 1066 (2019)

    Article  Google Scholar 

  10. Kingma, D.P., Ba, J.: Adam: a method for stochastic optimization (2017)

    Google Scholar 

  11. Li, X., Yang, J., Ma, J.: Recent developments of content-based image retrieval (CBIR). Neurocomputing 452, 675–689 (2021)

    Article  Google Scholar 

  12. Lin, C.Y.: Rouge: a package for automatic evaluation of summaries. In: Text summarization branches out, pp. 74–81 (2004)

    Google Scholar 

  13. Lin, T.-Y., et al.: Microsoft COCO: common objects in context. In: Fleet, D., Pajdla, T., Schiele, B., Tuytelaars, T. (eds.) ECCV 2014. LNCS, vol. 8693, pp. 740–755. Springer, Cham (2014). https://doi.org/10.1007/978-3-319-10602-1_48

    Chapter  Google Scholar 

  14. Ma, H., et al.: Ei-clip: entity-aware interventional contrastive learning for e-commerce cross-modal retrieval. In: CVPR, pp. 18051–18061 (2022)

    Google Scholar 

  15. Messina, N., et al.: Fine-grained visual textual alignment for cross-modal retrieval using transformer encoders. ACM Trans. Multimedia Comput. Commun. Appl. (TOMM), 17(4), 1–23 (2021)

    Google Scholar 

  16. Messina, N., Falchi, F., Esuli, A., Amato, G.: Transformer reasoning network for image-text matching and retrieval. In: 2020 25th International Conference on Pattern Recognition (ICPR), pp. 5222–5229. IEEE (2021)

    Google Scholar 

  17. Molina, G., et al.: A new content-based image retrieval system for SARS-CoV-2 computer-aided diagnosis. In: Su, R., Zhang, Y.-D., Liu, H. (eds.) MICAD 2021. LNEE, vol. 784, pp. 316–324. Springer, Singapore (2022). https://doi.org/10.1007/978-981-16-3880-0_33

    Chapter  Google Scholar 

  18. Anderson, P., Fernando, B., Johnson, M., Gould, S.: SPICE: semantic propositional image caption evaluation. In: Leibe, B., Matas, J., Sebe, N., Welling, M. (eds.) ECCV 2016. LNCS, vol. 9909, pp. 382–398. Springer, Cham (2016). https://doi.org/10.1007/978-3-319-46454-1_24

    Chapter  Google Scholar 

  19. Ren, R., et al.: Pair: leveraging passage-centric similarity relation for improving dense passage retrieval, pp. 2173–2183 (2021)

    Google Scholar 

  20. Schubert, E.: A triangle inequality for cosine similarity. In: Reyes, N., et al. (eds.) SISAP 2021. LNCS, vol. 13058, pp. 32–44. Springer, Cham (2021). https://doi.org/10.1007/978-3-030-89657-7_3

    Chapter  Google Scholar 

  21. Song, H.O., Xiang, Y., Jegelka, S., Savarese, S.: Deep metric learning via lifted structured feature embedding. In: IEEE CVPR, pp. 4004–4012 (2016)

    Google Scholar 

  22. Song, K., Tan, X., Qin, T., Lu, J., Liu, T.Y.: Mpnet: masked and permuted pre-training for language understanding. NIPS 33, 16857–16867 (2020)

    Google Scholar 

  23. Song, Y., Soleymani, M.: Polysemous visual-semantic embedding for cross-modal retrieval. In: CVPR, pp. 1979–1988 (2019)

    Google Scholar 

  24. Tan, M., Le, Q.V.: Efficientnetv2: smaller models and faster training. CoRR abs/2104.00298 (2021)

    Google Scholar 

  25. Tian, Y., et al.: Sosnet: second order similarity regularization for local descriptor learning, pp. 11008–11017 (2019)

    Google Scholar 

  26. Ng, T., Balntas, V., Y, Tian., Mikolajczyk, K.: Solar: Second-order loss and attention for image retrieval. ArXiv (2020)

    Google Scholar 

  27. Wang, Z., et al.: Adaptive margin based deep adversarial metric learning. In: IEEE BigDataSecurity/HPSC/IDS 2020, pp. 100–108 (2020)

    Google Scholar 

  28. Chen, W., Chen, X., Zhang, J., Huang, K.: Beyond triplet loss: a deep quadruplet network for person re-identification. In: IEEE CVPR, pp. 1320–1329 (2017)

    Google Scholar 

  29. Wu, Y., Wang, S., Huang, Q.: Online asymmetric similarity learning for cross-modal retrieval. In: IEEE CVPR, pp. 3984–3993 (2017)

    Google Scholar 

  30. Wu, Y., Wang, S., Huang, Q.: Online fast adaptive low-rank similarity learning for cross-modal retrieval. IEEE Trans. Multimedia 22(5), 1310–1322 (2020)

    Article  Google Scholar 

  31. Xuan, H., Stylianou, A., Liu, X., Pless, R.: Hard negative examples are hard, but useful. In: Vedaldi, A., Bischof, H., Brox, T., Frahm, J.-M. (eds.) ECCV 2020. LNCS, vol. 12359, pp. 126–142. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-58568-6_8

    Chapter  Google Scholar 

  32. Yang, J., et al.: Vision-language pre-training with triple contrastive learning. In: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 15650–15659 (2022)

    Google Scholar 

  33. Ye, M., et al.: Deep learning for person re-identification: a survey and outlook. IEEE Trans. Pattern Anal. Mach. Intell. 44(6), 2872–2893 (2021)

    Article  Google Scholar 

  34. Young, P., Lai, A., Hodosh, M., Hockenmaier, J.: From image descriptions to visual denotations: new similarity metrics for semantic inference over event descriptions. TACL 2, 67–78 (2014)

    Article  Google Scholar 

  35. Zhao, C., et al.: Deep fusion feature representation learning with hard mining center-triplet loss for person re-identification. IEEE Trans. Multimedia 22(12), 3180–3195 (2020)

    Article  Google Scholar 

  36. Zhou, T., et al.: Solving the apparent diversity-accuracy dilemma of recommender systems. PNAS 107, 4511–4515 (2010)

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Universidad Federico Santa María, Valparaíso, Chile

    Mario Mallea, Ricardo Nanculef & Mauricio Araya

Authors
  1. Mario Mallea
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Ricardo Nanculef
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Mauricio Araya
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Mario Mallea .

Editor information

Editors and Affiliations

  1. Waikato University, Hamilton, New Zealand

    Albert Bifet

  2. Aeronautics Institute of Technology, São José dos Campos, Brazil

    Ana Carolina Lorena

  3. University of Porto, Porto, Portugal

    Rita P. Ribeiro

  4. University of Porto, Porto, Portugal

    João Gama

  5. University of Coimbra, Coimbra, Portugal

    Pedro H. Abreu

Rights and permissions

Reprints and permissions

Copyright information

© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this paper

Check for updates. Verify currency and authenticity via CrossMark

Cite this paper

Mallea, M., Nanculef, R., Araya, M. (2023). Enhancing Intra-modal Similarity in a Cross-Modal Triplet Loss. In: Bifet, A., Lorena, A.C., Ribeiro, R.P., Gama, J., Abreu, P.H. (eds) Discovery Science. DS 2023. Lecture Notes in Computer Science(), vol 14276. Springer, Cham. https://doi.org/10.1007/978-3-031-45275-8_17

Download citation

  • DOI: https://doi.org/10.1007/978-3-031-45275-8_17

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-031-45274-1

  • Online ISBN: 978-3-031-45275-8

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics

Search

Navigation