Abstract
Although generalized zero-shot learning (GZSL) has achieved success in recognizing images of unseen classes, most previous studies focused on feature projection from one domain to another, neglecting the importance of semantic descriptions. In this paper, we propose auxiliary-features via GAN(Af-GAN) to deal with the semantic loss problem. Auxiliary-features contain both real features of seen classes and instructive-features mapped by attributes. For the seen classes, we deploy the auxiliary-features to train the generator and regularize the synthesized samples to be close to auxiliary-features. For the unseen classes, we take the instructive-feature mapped by attributes to synthesize unseen class samples for training the final classifier. We construct the constraint between real visual features and instructive-features, and reduce the dependence on the class attributes. Considering that features synthesized from a set of similar attributes overlap each other in visual space, we combine Cosine similarity and Euclidean distance to constrain the distribution of synthesized features. Our method outperforms state-of-the-art methods on four benchmark datasets and also surpasses prior work by a large margin in generalized zero-shot learning.
This research was supported by the Fundamental Research Funds for the Central Universities, and the Research Funds of Renmin University of China (20XNA031).The computer resources were provided by Public Computing Cloud Platform of Renmin University of China.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Akata, Z., Perronnin, F., Harchaoui, Z., et al.: Label-embedding for image classification. IEEE Trans. Pattern Anal. Mach. Intell. 38, 1428–1435 (2016)
Lampert, C.H., Nickisch, H., Harmeling, S.: Learning to detect unseen object classes by between class attribute transfer. In: IEEE Conference on Computer Vision and Pattern Recognition (2009)
Liu, J., Zhang, Z., Yang, G.: Cross-class generative network for zero-shot learning. Inf. Sci. 555, 147–163 (2021)
Liu, L., Zhou, T., Long, G., Jiang, J., Dong, X., Zhang, C.: Isometric propagation network for generalized zero-shot learning. In: International Conference on Learning Representations (2021)
Bo, L., Dong, Q., Hu, Z.: Hardness sampling for self-training based transductive zero-shot learning. In: Computer Vision and Pattern Recognition, pp. 16499–16508 (2021)
Xie, C., Xiang, H., Zeng, T., Yang, Y., Yu, B., Liu, Q.: Cross knowledge-based generative zero-shot learning approach with taxonomy regularization. Neural Netw. 139, 168–178 (2021)
Ye, M., Guo, Y.: Progressive ensemble networks for zero-shot recognition. In: IEEE Conference on Computer Vision and Pattern (2019)
Kampffmeyer, M., Chen, Y., Liang, X., et al.: Rethinking knowledge graph propagation for zero-shot learning. In: IEEE Conference on Computer Vision and Pattern (2019)
Schonfeld, E., Ebrahimi, S., Sinha, S., et al.: Generalized zero- and few-shot learning via aligned variational autoencoders. In: IEEE Conference on Computer Vision and Pattern (2019)
Changpinyo, S., Chao, W.L., Sha, F.: Predicting visual exemplars of unseen classes for zero-shot learning. In: IEEE Conference on Computer Vision and Pattern (2016)
Kodirov, E., Xiang, T., Gong, S.: Semantic autoencoder for zero-shot learning. In: IEEE Conference on Computer Vision and Pattern Recognition (2017)
Zhu, P., Wang, H., Saligrama, V.: Generalized zero-shot recognition based on visually semantic embedding. In: IEEE Conference on Computer Vision and Pattern (2020)
Mandal, D., Narayan, S., Dwivedi, S., et al.: Out-of-distribution detection for generalized zero-shot action recognition. In: IEEE Conference on Computer Vision and Pattern (2019)
Huynh, D., Elhamifar, E.: Fine-grained generalized zero-shot learning via dense attribute-based attention. In: IEEE Conference on Computer Vision and Pattern (2019)
Han, Z., Fu, Z., Yang, J.: Learning the redundancy-free features for generalized zero-shot object recognition. In: IEEE Conference on Computer Vision and Pattern (2020)
Mirza, M., Osindero, S.: Conditional generative adversarial nets. arXiv preprint arXiv:1411.1784 (2014)
Zhang, Z., Saligrama, V.: Zero-shot learning via semantic similarity embedding. In: IEEE Conference on Computer Vision and Pattern (2015)
Lampert, C.H., Nickisch, H., Harmeling, S.: Attribute-based classification for zero-shot visual object categorization. IEEE Transactions on Pattern Analysis Machine Intelligence. 36, 435–453 (2014)
Li, J., Jin, M., Lu, K., et al.: Leveraging the invariant side of generative zero-shot learning. In: IEEE Conference on Computer Vision and Pattern (2019)
Xian, Y., Lorenz, T., Schiele, B., Akata, Z.: Feature generating networks for zero-shot learning. In: IEEE Conference on Computer Vision and Pattern (2018)
Keshari, R., Singh, R., Vatsa, M.: Generalized zero-shot learning via over-complete distribution. In: IEEE Conference on Computer Vision and Pattern (2020)
Romera-Paredes, B., Torr, P.H.S.: An embarrassingly simple approach to zero-shot learning. In: Proceedings of the 32nd International Conference on Machine Learning (2015)
Norouzi, M., et al.:. Zero-shot learning by convex combination of semantic embeddings. In: International Conference on Learning Representations (2014)
Paul, A., Krishnan, N.C., Munjal, P.: Semantically aligned bias reducing zero shot learning. In: IEEE Conference on Computer Vision and Pattern (2019)
Deng, J., Guo, J., Xue, N., Zafeiriou, S.: ArcFace: additive angular margin loss for deep face recognition. In: Computer Vision and Pattern Recognition, pp. 4690–4699 (2019)
He, G., Li, F., Wang, Q., Bai, Z., Xu, Y.: A hierarchical sampling based triplet network for fine-grained image classification. Pattern Recogn. 115, 107889 (2021)
Ji, Z., Sun, Y., Yu, Y., Pang, Y., Han, J.: Attribute-guided network for cross-modal zero-shot hashing. IEEE Trans. Neural Netw. 31(1), 321–330 (2020)
Akata, Z., et al.: Evaluation of output embeddings for fine-grained image classification. In: IEEE Conference on Computer Vision and Pattern Recognition (2015)
Akata, Z., et al.: Label-embedding for image classification. In: IEEE Transactions on Pattern Analysis and Machine Intelligence (2016)
Gao, R., Hou, X., Qin, J., Liu, L., Zhu, F., Zhang, Z.: A joint generative model for zero-shot learning. In: Leal-Taixé, L., Roth, S. (eds.) ECCV 2018. LNCS, vol. 11132, pp. 631–646. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-11018-5_50
Kumar Verma, V., Arora, G., Mishra, A., Rai, P.: Generalized zero-shot learning via synthesized examples. In: IEEE Conference on Computer Vision and Pattern (2018)
Zhang, H., Long, Y., Liu, L., Shao, L.: Adversarial unseen visual feature synthesis for zero-shot learning. Neurocomputing 329, 12–20 (2019)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Sun, W., Yang, G. (2023). Generative Generalized Zero-Shot Learning Based on Auxiliary-Features. In: Tanveer, M., Agarwal, S., Ozawa, S., Ekbal, A., Jatowt, A. (eds) Neural Information Processing. ICONIP 2022. Communications in Computer and Information Science, vol 1791. Springer, Singapore. https://doi.org/10.1007/978-981-99-1639-9_44
Download citation
DOI: https://doi.org/10.1007/978-981-99-1639-9_44
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-99-1638-2
Online ISBN: 978-981-99-1639-9
eBook Packages: Computer ScienceComputer Science (R0)