About

Changhu Wang is a Leader Researcher with Microsoft Research, Beijing, China. His current research interests include multimedia analysis and retrieval, computer vision, and machine learning. In particular, he is interested in applying the techniques from these areas to a broad range of multimedia and vision applications, such as sketch visual analysis, image understanding, multimedia search, and deep learning.

Changhu has authored or co-authored over 50 papers in high-quality journals and conferences, such as CVPR, ICCV, ECCV, ACM Multimedia, SIGIR, WWW, ICML, IJCAI, AAAI, TIP, TMM. He has shipped multiple technologies to Microsoft products, such as Bing, Office, XiaoIce, etc. He holds 7 U.S. granted patents and 5 in pending. He also served as program committee member and reviewer for 20+ high-quality conferences and journals, such as TPAMI, TIP, TMM, TCSVT, TKDE, TOMCCAP, TIST, TIIS, WWW, SIGIR, ACM MM, ICCV, SIGGRAPH. He is a Senior Member of the IEEE and the ACM. He built one of the first million-level sketch-based image search systems, as well as the first billion-level system in the world. He also built one of the first general sketch recognition systems. He was awarded Microsoft Gold Star Award in 2011, which is a highly selective award in the recognition of his important contribution to Microsoft’s success. He received a Best Demo Award in ACM International Conference on Multimedia 2010 in the recognition of MindFinder system. He was awarded Mi-crosoft Fellowship in 2007.

Changhu received B.E. and Ph.D. degrees from the University of Science and Technology of China, in 2004 and 2009, respectively. He worked as a research engineer at the department of Electrical and Computer Engineering in National University of Singapore in 2008. He is a guest professor of Xi’an University of Posts and Telecommunications.

Projects

Network Morphism

Established: September 1, 2015

We present a systematic study on how to morph a well-trained neural network to a new one so that its network function can be completely preserved. We define this as network morphism in this research. After morphing a parent network,…

Image Captioning

Established: January 25, 2016

We study the problem of image captioning, i.e., automatically describing an image by a sentence. we formulate the problem of image captioning as a multimodal translation task. Analogous to machine translation, we present a sequence-to-sequence recurrent neural networks (RNN) model…

Food Recognition

Established: January 25, 2016

  We study the problem of food image recognition via deep learning techniques. Our goal is to develop a robust service to recognize thousands of popular Asia and Western food. Several prototypes have been developed to support diverse applications. The…

MindFinder: Finding Images by Sketching

Established: August 12, 2009

Sketch-based image search is a well-known and difficult problem, in which little progress has been made in the past decade in developing a large-scale and practical sketch-based search engine. We have revisited this problem and…

Publications

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Projects

Other

  1. Invited talk, “Sketch-based Image Retrieval and Related Applications”, Qinhuangdao Campus, Northeastern University, Qinhuangdao, China, 07/2015.
  2. Invited talk, “Sketch-based Image Search and Related Applications,” Xi’an University of Posts and Telecommunications, Beijing, China, 10/2014.
  3. Course talk, “Visual Information Retrieval,” Renmin University, Beijing, China, 05/2014.
  4. Invited talk, “Sketch-based Image Search and Related Applications,” Cloud4Science summer school, CAS, Beijing, China, 08/2013.
  5. Invited talk, “MindFinder: Sketch-based Search, Recognition, and Parsing,” Peking University, Beijing, China, 03/2013.
  6. Invited talk, “MindFinder: Sketch-based Search, Recognition, and Parsing,” Idiap Research Institute, Martigny, Switzerland, 11/2012.
  7. Invited talk, “Edgel Index for Sketch-based Search, Recognition, and Parsing,” Tsinghua University, Beijing, China, 06/2012.
  8. Invited talk, “Sketch-based Image Search: Technologies, Applications, and Challenges,” University of Science and Technology of China, Hefei, China, 05/2012.
  9. Invited talk, “MindFinder: Finding Images by Sketching,” Shanghai Jiao Tong University, Beijing, China, 04/2011.
  10. Invited talk, “Several works on Web Image Search,” School of Computing, National University of Singapore, Singapore, 06/2008.
  11. Invited talk, “Several works on Web Image Search,” Tsinghua University, Beijing, China, 12/2007.

Awards

  1. Microsoft Gold Star Award, in the recognition of the MindFinder system, as well as the contribution to Microsoft, 2011.
  2. The Best Demo Award in ACM Multimedia conference, in the recognition of the MindFinder system, 2010.
  3. Microsoft Fellowship, 2007.

Patents

  1. Changhu Wang, Lei Zhang, Zhenbang Sun. Sketch Segmentation. United States Patent Application Publication. Patent No. US 9,183,467 B2. Date of Patent: Nov. 10, 2015. (Granted)
  2. Changhu Wang, Lei Zhang, Zhenbang Sun. Hand-Drawn Sketch Recognition. United States Patent Application Publication.. Patent No. US 9,147,125 B2. Date of Patent: Sep. 29, 2015. (Granted)
  3. Rong Xiao, Qiang Hao, Changhu Wang, Rui Cai, Lei Zhang. Snippet Extraction and Ranking. United States Patent Application Publication. Patent No. US 8,945,425 B2. Date of Patent: Feb. 10, 2015. (Granted)
  4. Changhu Wang, Zhiwei Li, Lei Zhang. Image Retrieval using Spatial Bag-of-Features. United States Patent. Patent No. US 8,849,030 B2. Date of Patent: Sep. 30, 2014. (Granted)
  5. Rui Cai, Qiang Hao, Changhu Wang, Rong Xiao, Lei Zhang. Mining Topic-Related Aspects from User Generated Content. United States Patent. Patent No. US 8,458,115 B2. Date of Patent: June 4, 2013. (Granted)
  6. Changhu Wang, Zhiwei Li, Lei Zhang. Image search by interactive sketching and tagging. United States Patent. Patent No. US 8,447,752 B2. Date of Patent: May 21, 2013. (Granted)
  7. Feng Jing, Lei Zhang, Mingjing Li, Wei-Ying Ma, Changhu Wang. Generating clusters of images for search results. United States Patent. Patent No. US 7,725,451 B2. Date of Patent: May 25, 2010. (Granted)
  8. Changhu Wang, Jiang-Ming Yang, Lei Zhang, Xin Lu. Travel Route Planning Using Geo-Tagged Photographs. United States Patent Application Publication. Pub. No. US 2012/0084000 A1. Pub. Date: Apr. 5, 2012. (Pending)
  9. Changhu Wang, Zhiwei Li, Lei Zhang. Sketch-based Image Search. United States Patent Application Publication. Pub. No. US 2012/0054177 A1. Pub. Date: Mar. 1, 2012. (Pending)
  10. Changhu Wang, Avneesh Sud, Lei Zhang, Xinghai Sun. United States Patent Application. MS# 338922.01. Filed Date: Jun. 14, 2013. (Filed)
  11. Multi-signal based shopping cart content recognition in brick-and-mortar retail stores. (Filed)
  12. Network Morphism. (Filed)