20160727-24 Xin Lu:Towards Automatic Prediction of Image Aesthetics

报告嘉宾2：Xin Lu（Adobe, USA）

报告时间：2016年7月27日（星期三）晚21:00（北京时间）

报告题目：Towards Automatic Prediction of Image Aesthetics 

主持人： 汪张扬（Texas A&M University, USA）

报告摘要：

Effective visual features are essential for computational aesthetic quality rating systems. Existing methods used machine learning and statistical modeling techniques on handcrafted features or generic image descriptors. A recently-published large-scale dataset, the AVA dataset, has further empowered machine learning based approaches. In this talk, I will present an image aesthetics prediction approach that incorporates heterogeneous inputs generated from the image, which include a global view and a local view, and unifies the feature learning and classier training using a double-column deep convolutional neural network. Meanwhile, a deep multi-patch aggregation network training approach will also be introduced, which allows us to better capture fine-grained details from high-resolution images. The effectiveness of this approach have been demonstrated on the problems of image style recognition, aesthetic quality categorization, and image quality estimation.

参考文献：

[1] Xin Lu, Zhe Lin, Xiaohui Shen, Radomir Mech and James Z. Wang, ``Deep Multi-Patch Aggregation Network for Image Style, Aesthetics, and Quality Estimation,'' Proceedings of the International Conference on Computer Vision, pp. 990-998, Santiago, Chile, IEEE, 2015.

[2] Xin Lu, Zhe Lin, Hailin Jin, Jianchao Yang and James Z. Wang, ``Rating Pictorial Aesthetics using Deep Learning,'' IEEE Transactions on Multimedia, vol. 17, no. 11, pp, 2021-2034, 2015.

[3] Xin Lu, Zhe Lin, Hailin Jin, Jianchao Yang and James Z. Wang, ``RAPID: Rating Pictorial Aesthetics using Deep Learning,'' Proceedings of the ACM Multimedia Conference, pp. 457-466,

Orlando, Florida, ACM, November 2014.

报告人简介：

Xin Lu received her B.E. and B.A. degree in Electronic Information and Engineering and English, the M.E. Degree in Signal and Information Processing from Tianjin University, China. She received her Ph.D. degree from the College of Information Sciences and Technology, The Pennsylvania State University, University Park, USA. She worked as a research intern at Microsoft Research Asia in 2009-2010 and at Adobe in the summers of 2012, 2013, and 2014. She started working at Adobe as a Research Scientist since Aug, 2015. Her research interests include computer vision and multimedia analysis, deep learning, image processing, Web search and mining, and user-generated content analysis.