Chen Fang
6211 Sudikoff Lab Phone: (603) 277-0542
Dartmouth College Email:
Hanover NH, 03755 Homepage: chenfang/
Education
Ph.D. Computer Science, Dartmouth College Sept. 2010 - Expected Jun. 2015
B.E. Software Engineering, Hunan University Sept. 2006 - Jun. 2010
Expertise
Programming Skill
: Proficient in C, C++, Python (Cython) and Matlab. Some experience with Java.
Software
: Eclipse, Visual Studio, Latex. GNU/Linux systems, Windows Systems.
Experience
Research Intern, Adobe Research, Imagination Lab, San Jose, CA Jun. 2014 - Sept. 2014
Project
: Image Feature Learning from Large Scale Social Media Data
* Collaborated with research and product teams on learning image features for better visual similarity for a
popular social media website.
* Developed a weakly supervised deep learning method to learn image features from large scale noisy user
behavior data.
* Wrote highly efficient code to process and run collaborative filtering on hundreds of million of records.
* Achieved best accuracy on various image classification benchmarks, and submitted a paper to a top tier con-
ference.
Research Intern, Adobe Research, Imagination Lab, San Jose, CA Jun. 2013 - Sept. 2013
Project
: Smart Image Cropping
* Worked with research and product teams to develop an automatic image cropping system to improve the com-
position quality of photographs.
* Developed a novel machine learning based image cropping algorithm, which significantly outperformed com-
peting methods. The developed algorithm was integrated into Photoshop Lightroom and Mix.
* Published a paper in a top tier conference. Filed three US patents.
Software
Engineer Intern, A9.com/Amazon, Visual Search Team, Palo Alto, CA Jun. 2012 - Sept. 2012
Project
: 3D Modeling of Online Retail Products
* Worked with computer vision engineers to develop a framework to generate 3D models for Amazon products.
* Refined the noisy 3D model by fusing Kinect data with high resolution imagery data from SLR camera.
Research Assistant, Dartmouth College, Hanover, NH Sept. 2010 - present
Project
: Ph.D. research on computer vision and machine learning
* Worked on various problems in computer vision and machine learning, including large scale image retrieval,
image recognition, feature learning, metric learning and structural learning. Published papers in top tier con-
ferences.
Chen Fang 2
Publications
* Chen Fang, Hailin Jin, Jianchao Yang, Zhe Lin. Collaborative Feature Learning from Social Media.
IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2015.
* Chen Fang, Dan Rockmore. Multi-Task Metric Learning on Network Data.
Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD) 2015.
* Chen Fang, Zhe Lin, Radomir Mech, Xiaohui Shen. Automatic Image Cropping using Visual Composition,
Boundary Simplicity and Content Preservation Models.
ACM Multimedia (MM) 2014.
(The algorithm developed in this paper has been selected to ship with Photoshop Lightroom and Mix, enabling the
new auto cropping feature for the first time.)
* Chen Fang, Ye Xu, Dan Rockmore. Unbiased Metric Learning: On the Utilization of Multiple Datasets and Web
Images for Softening Bias.
International Conference on Computer Vision (ICCV) 2013.
* Chen Fang, Lorenzo Torresani. Measuring Image Distances via Embedding in a Semantic Manifold.
European Conference on Computer Vision (ECCV) 2012.
* Chen Fang*, Mohammad Rastegari*, Lorenzo Torresani. Scalable Object-Class Retrieval with Approximate and
Top-k Ranking. (* means equal contribution)
International Conference on Computer Vision (ICCV) 2011.
* Chen Fang, Jin Zhang, Shangwu Zhu, Guang Li, Rulong Wang. Analysis of Texture Images Generated by Olfac-
tory System Bionic Model.
IEEE International Symposium on Neural Networks (ISNN) 2010.
* Chen Jian-zhou, Zhang Jin, Chen Fang, Zhao Li-juan, Hans Liljenstrom. On Color Texture Generating Based on
Simplified KIII Model.
IEEE/ACIS International Conference on Computer and Information Science (ICIS) 2009.
Patents
* Zhe Lin, Radomir Mech, Xiaohui Shen, Chen Fang. Category Histogram: A High Level Image Representation via
Image Search and Localization. U.S. patent 81170286. (Filed on Feb 2014)
* Zhe Lin, Radomir Mech, Xiaohui Shen, Chen Fang. Smart Image Cropping by Combining Composition, Content
Preservation, and Crop Boundary Simplicity Cues. U.S. patent 81169931. (Filed on Jan 2014)
* Zhe Lin, Radomir Mech, Xiaohui Shen, Chen Fang. Boundary Simplicity Model for Automatic Image Cropping.
U.S. patent 81172242. (Filed on Jan 2014)
Teaching Experience
Dartmouth College, Teaching Assistant for Machine Learning, Computer Vision, Graphics. 2011 - 2013
Software
VLG extractor: Efficient implementation to extract the image descriptors proposed in [Torresani et al., ECCV
2010], [Bergamo et al., NIPS 2011] and [Bergamo and Torresani, CVPR 2012]. Written in C/C++ using OpenCV
and BLAS. Binaries available for Linux 64bit, Windows 32bit, and MacOSX 10.6.
Dartmouth Fellowship
: 2010 - 2015
National Scholarship of China (top 2%): 2010