Abstract
Predictor combination aims to improve a (target) predictor of a learning task based on the (reference) predictors of potentially relevant tasks, without having access to the internals of individual predictors. We present a new predictor combination algorithm that improves the target by i) measuring the relevance of references based on their capabilities in predicting the target, and ii) strengthening such estimated relevance. Unlike existing predictor combination approaches that only exploit pairwise relationships between the target and each reference, and thereby ignore potentially useful dependence among references, our algorithm jointly assesses the relevance of all references by adopting a Bayesian framework. This also offers a rigorous way to automatically select only relevant references. Based on experiments on seven real-world datasets from visual attribute ranking and multi-class classification scenarios, we demonstrate that our algorithm offers a significant performance gain and broadens the application range of existing predictor combination approaches.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Here, the term ‘linear’ signifies the capability of \(\mathcal {K}_\text {L}\) to capture the linear dependence of references, independent of \(\mathcal {K}_\text {L}\) being a linear operator as well.
References
Argyriou, A., Evgeniou, T., Pontil, M.: Convex multi-task feature learning. Mach. Learn. 73(3), 243–272 (2008). https://doi.org/10.1007/s10994-007-5040-8
Chen, L., Zhang, Q., Li, B.: Predicting multiple attributes via relative multi-task learning. In: CVPR, pp. 1027–1034 (2014)
Everingham, M., Eslami, S.M.A., Van Gool, L., Williams, C.K.I., Winn, J., Zisserman, A.: The Pascal visual object classes challenge: a retrospective. Int. J. Comput. Vis. 111(1), 98–136 (2015). https://doi.org/10.1007/s11263-014-0733-5
Evgeniou, T., Micchelli, C.A., Pontil, M.: Learning multiple tasks with kernel methods. JMLR 6, 615–637 (2005)
Farhadi, A., Endres, I., Hoiem, D., Forsyth, D.: Describing objects by their attributes. In: CVPR, pp. 1778–1785 (2009)
Gong, P., Ye, J., Zhang, C.: Robust multi-task feature learning. In: KDD, pp. 895–903 (2012)
He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: CVPR, pp. 770–778 (2016)
Hein, M., Maier, M.: Manifold denoising. In: NIPS, pp. 561–568 (2007)
Joachims, T.: Optimizing search engines using clickthrough data. In: KDD, pp. 133–142 (2002)
Kim, K.I., Chang, H.J.: Joint manifold diffusion for combining predictions on decoupled observations. In: CVPR, pp. 7549–7557 (2019)
Kim, K.I., Tompkin, J., Richardt, C.: Predictor combination at test time. In: ICCV, pp. 3553–3561 (2017)
Kovashka, A., Parikh, D., Grauman, K.: WhittleSearch: image search with relative attribute feedback. In: CVPR, pp. 2973–2980 (2012)
Mejjati, Y.A., Cosker, D., Kim, K.I.: Multi-task learning by maximizing statistical dependence. In: CVPR, pp. 3465–3473 (2018)
Meng, Z., Adluru, N., Kim, H.J., Fung, G., Singh, V.: Efficient relative attribute learning using graph neural networks. In: Ferrari, V., Hebert, M., Sminchisescu, C., Weiss, Y. (eds.) Computer Vision – ECCV 2018. LNCS, vol. 11218, pp. 575–590. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-01264-9_34
Oliva, A., Torralba, A.: Modeling the shape of the scene: a holistic representation of the spatial envelope. Int. J. Comput. Vis. 42(3), 145–175 (2001). https://doi.org/10.1023/A:1011139631724
Pan, S.J., Yang, Q.: A survey on transfer learning. IEEE Trans. Knowl. Data Eng. 22(10), 1345–1359 (2010)
Parikh, D., Grauman, K.: Relative attributes. In: ICCV, pp. 503–510 (2011)
Passos, A., Rai, P., Wainer, J., Daumé III, H.: Flexible modeling of latent task structures in multitask learning. In: ICML, pp. 1103–1110 (2012)
Rasmussen, C.E., Williams, C.K.I.: Gaussian Processes for Machine Learning. MIT Press, Cambridge (2006)
Seeger, M., Williams, C.K.I., Lawrence, N.D.: Fast forward selection to speed up sparse Gaussian process regression. In: International Workshop on Artificial Intelligence and Statistics (2003)
Snelson, E., Ghahramani, Z.: Sparse Gaussian processes using pseudo-inputs. In: NIPS (2006)
Wah, C., Branson, S., Welinder, P., Perona, P., Belongie, S.: The Caltech-UCSD Birds-200-2011 Dataset. Technical report CNS-TR-2011-001, California Institute of Technology (2011)
Xian, Y., Lampert, C.H., Schiele, B., Akata, Z.: Zero-shot learning - a comprehensive evaluation of the good, the bad and the ugly. IEEE TPAMI 41(9), 2251–2265 (2019)
Yang, X., Zhang, T., Xu, C., Yan, S., Hossain, M.S., Ghoneim, A.: Deep relative attributes. IEEE T-MM 18(9), 1832–1842 (2016)
Yu, A., Grauman, K.: Fine-grained visual comparisons with local learning. In: CVPR, pp. 192–199 (2014)
Zamir, A.R., Sax, A., Shen, W., Guibas, L., Malik, J., Savarese, S.: Taskonomy: disentangling task transfer learning. In: CVPR, pp. 3712–3722 (2018)
Zhou, X., Belkin, M., Srebro, N.: An iterated graph Laplacian approach for ranking on manifolds. In: KDD, pp. 877–885 (2011)
Acknowledgements
This work was supported by UNIST’s 2020 Research Fund (1.200033.01), National Research Foundation of Korea (NRF) grant NRF-2019R1F1A1061603, and Institute of Information & Communications Technology Planning & Evaluation (IITP) grant (No. 20200013360011001, Artificial Intelligence Graduate School support (UNIST)) funded by the Korean government (MSIT).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
1 Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Kim, K.I., Richardt, C., Chang, H.J. (2020). Combining Task Predictors via Enhancing Joint Predictability. In: Vedaldi, A., Bischof, H., Brox, T., Frahm, JM. (eds) Computer Vision – ECCV 2020. ECCV 2020. Lecture Notes in Computer Science(), vol 12361. Springer, Cham. https://doi.org/10.1007/978-3-030-58517-4_26
Download citation
DOI: https://doi.org/10.1007/978-3-030-58517-4_26
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-58516-7
Online ISBN: 978-3-030-58517-4
eBook Packages: Computer ScienceComputer Science (R0)