59-05-032 Proceeding

83 Proceedings of the Princess Maha Chakri Sirindhorn Congress method, CNN models and a state-of-the-art method. S-CNN refers to trained attribute-specific models of CNN. CF refers to the combined features model with no pose baseline [4], while CRF refers to the previous state-of-the-art method proposed by [5]. Table 4 The mean average accuracy (%) of attribute prediction on the Clothing dataset [23]. Method Accuracy S-CNN [10] 90.43 CF [4] 80.48 CRF [5] 83.95 2) Related Applications : A direct application that can leverage the learned semantic attribute representations, is to train a ranking algorithm to predict which dress images are more likable than others on our collected datasets. Acknowledgement This research was carried out at the Rapid-Rich Object Search (ROSE) Lab at the Nanyang Technological University, Singapore. The ROSE Lab is supported by the National Research Foundation, Prime Minister’s Office, Singapore, under its IDM Futures Funding Initiative and administered by the Interactive and Digital Media Programme Office. We also gratefully acknowledge the support of NVIDIA Corporation for their donation of Tesla K40 GPUs used for our research at the ROSE Lab. References Si Liu, Luoqi Liu, and Shuicheng Yan, “Fashionanalysis: Current techniques and future directions,” MultiMedia, IEEE, vol. 21, no. 2, pp. 72–79, 2014. Si Liu, Jiashi Feng, Zheng Song, Tianzhu Zhang, Hanqing Lu, Changsheng Xu, and Shuicheng Yan, “Hi, magic closet, tell me what to wear!,” in Proceedings of the 20 th ACM international conference on Multimedia . ACM, 2012, pp. 619–628. Basela Hasan and David Hogg, “Segmentation using deformable spatial priors with application to clothing.” in BMVC, 2010 , pp. 1–11. A. Gallagher and T. Chen. “Clothing segmentation for recognizing people”, in Computer Vision and Pattern Recognition, 2008. CVPR 2008. IEEE Conference on. IEEE, 2008, pp. 1-8. H. Chen, A. Gallagher, and B. Girod. “Describing clothing by semantic attributes,” in Computer Vision–ECCV 2012 , pp. 609–623. Springer, 2012. http://caffe.berkeleyvision.org/ http://www.vlfeat.org/matconvnet/ https://github.com/NVIDIA/DIGITS