Abstract
A new SVM training algorithm is proposed in the paper to improve the validity and efficiency of image annotation. These annotation tasks are related to one another due to the correlation among the labels. The model will implicitly learn a linear output kernel during training. Simulation results show that compared with independent SVMs training, Joint SVM improves classification accuracy and efficiency substantially.
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
Similar content being viewed by others
References
Lavrenko V, Manmatha R, Jeon J (2004) A model for learning the semantics of pictures. In: Proceeding of advances in neural information processing systems, pp 553–560
Blei DM, Jordan MI (2003) Modeling annotated data. In: Proceeding of the 26th annual international ACM SIGIR conference on research and development information retrieval, pp 127–134
Feng SL, Manmatha R, Lavrenko V (2004) Multiple Bernoulli relevance models for image and video annotation. In: Proceeding of the 2nd IEEE computer society conference on computer vision and pattern recognition, vol 2. IEEE, pp II-1002–II-1009
Guillaumin M, Mensink T, Verbeek J et al (2009) TagProp: discriminative metric learning in nearest neighbor models for image auto-annotation. In: Proceeding of IEEE, international conference on computer vision. IEEE, pp 309–316
Makadia A, Pavlovic V, Kumar S (2010) Baselines for image annotation. Int J Comput Vis 90(1):88–105
Akgül YS, Kiliç MM. Ship location estimation from radar and optic images using metric learning. In: Proceeding of 26th signal processing and communications applications conference (SIU), Izmir, Turkey, pp 1–4
Nawaz S, Calefati A, Ahmed N et al (2018) Hand written characters recognition via Deep metric learning. In: Proceedings of 13th IAPR international workshop on document analysis systems (DAS), Vienna, Austria, pp 417–422
Wahlberg F (2018) Gaussian process classification as metric learning for forensic writer identification. In: Proceedings of 13th IAPR international workshop on document analysis systems (DAS), Vienna, Austria, pp 175–180
Elisseeff A, Weston J (2001) A kernel method for multi-labelled classification. In: Proceeding of international conference on neural information processing systems: natural and synthetic. MIT Press, pp 681–687
Guo Y, Schuurmans D (2013) Multi-label classification with output kernels. In: Proceeding of joint European conference on machine learning and knowledge discovery in databases. Springer, Berlin, Heidelberg, pp 417–432
Pan L, Li HC, Sun YJ et al (2018) Hyperspectral image reconstruction by latent low-rank representation for classification [J]. IEEE Geosci Remote Sens Lett 15(9):1422–1426
Hariharan B, Vishwanathan et al (2012) Efficient max-margin multi-label classification with applications to zero-shot learning. Mach Learn 88(1–2):127–155
Koda S, Zeggada A, Melgani F et al (2018) Spatial and structured SVM for multilabel image classification [J]. IEEE Trans Geosci Remote Sens 56(10):5948–5960
Argyriou A, Evgeniou T, Pontil M (2008) Convex multi-task feature learning [J]. Mach Learn 73(3):243–272
Dinuzzo F, Cheng SO, Gehler PV et al (2012) Learning output kernels with block coordinate descent [C]. In: Proceeding of international conference on machine learning, ICML 2011. Bellevue, Washington, USA, DBLP, pp 49–56
Zhang Y, Yeung DY (2013) Multilabel relationship learning [J]. ACM Trans Knowl Discov Data (TKDD) 7(2):7
Acknowledgements
This work is supported by the National High Technology Research and Development Program of China (2012AA120802), National Natural Science Foundation of China (61771186), Postdoctoral Research Project of Heilongjiang Province (LBH-Q15121), Undergraduate University Project of Young Scientist Creative Talent of Heilongjiang Province (UNPYSCT-2017125), Postgraduate Innovative Research Project of Heilongjiang University (NO. YJSCX2019-059HLJU), National Natural Science Foundation of China (61571162), Heilongjiang Province Natural Science Foundation (F2016019).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Feng, P. et al. (2020). Research on an Improved SVM Training Algorithm. In: Liang, Q., Wang, W., Liu, X., Na, Z., Jia, M., Zhang, B. (eds) Communications, Signal Processing, and Systems. CSPS 2019. Lecture Notes in Electrical Engineering, vol 571. Springer, Singapore. https://doi.org/10.1007/978-981-13-9409-6_201
Download citation
DOI: https://doi.org/10.1007/978-981-13-9409-6_201
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-13-9408-9
Online ISBN: 978-981-13-9409-6
eBook Packages: EngineeringEngineering (R0)