Computer Science & Engineering
Washington University in St. Louis 1 Brookings Dr. CB 1045, Saint Louis, MO 63130.
E-mail: ayan [at] wustl [dot] edu Office: Jolley 205
I work on problems in computer vision, computational photography, and machine learning—dealing with the design of accurate and efficient algorithms for visual inference, and of new kinds of high-capability sensors and cameras. I seek solutions to these problems by considering the physics of image formation, and the statistical structure of natural images and scenes.
[Google Scholar] [GitHub] [arXiv] [CV] news
Sep 2017 Joining WashU as Assistant Professor in CSE. Jun 2017 Work on learning beacon placement accepted to appear at IROS 2017. May 2017 New paper on stable GAN training posted to arXiv. May 2017 Will be serving as area chair for 3DV 2017 and CVPR 2018. students
I am looking to recruit talented PhD students. If you are interested, please apply directly to the
CSE PhD program. If you are an existing graduate or under-graduate student at WashU and would like to work with me, feel free to send me an e-mail.
research Stabilizing GAN Training with Multiple Random Projections arXiv 2017
Behnam Neyshabur, Srinadh Bhojanapalli, Ayan Chakrabarti
Paper: arXiv URL: Project website Jointly Optimizing Placement and Inference for Beacon-based Localization IROS 2017
Charles Schaff, David Yunis, Ayan Chakrabarti, Matthew R. Walter
Paper: arXiv URL: Project website Learning Sensor Multiplexing Design through Back-propagation NIPS 2016
Paper: arXiv URL: Project website Depth from a Single Image by Harmonizing Overcomplete Local Network Predictions NIPS 2016
Ayan Chakrabarti, Jingyu Shao, Gregory Shakhnarovich
Paper: arXiv URL: Project website Single-image RGB Photometric Stereo With Spatially-varying Albedo 3DV 2016 Oral
Ayan Chakrabarti, Kalyan Sunkavalli
Paper: arXiv URL: Project website A Neural Approach to Blind Motion Deblurring ECCV 2016
Paper: arXiv URL: Project website Color Constancy by Learning to Predict Chromaticity from Luminance NIPS 2015 Spotlight
Paper: arXiv (w/ supp. section) / NIPS URL: Project website Low-level Vision by Consensus in a Spatial Hierarchy of Regions CVPR 2015
Ayan Chakrabarti, Ying Xiong, Steven J. Gortler, Todd Zickler
Paper: CV-F / arXiv URL: Project website From Shading to Local Shape PAMI, Jan 2015
Ying Xiong, Ayan Chakrabarti, Ronen Basri, Steven J. Gortler, David W. Jacobs, Todd Zickler
Paper: IEEE Xplore / arXiv URL: Project website Modeling Radiometric Uncertainty for Vision with Tone-mapped Color Images PAMI, Nov 2014
Ayan Chakrabarti, Ying Xiong, Baochen Sun, Trevor Darrell, Daniel Scharstein, Todd Zickler, Kate Saenko
Paper: IEEE Xplore / arXiv URL: Project website Color Constancy with Spatio-Spectral Statistics PAMI, Aug 2012
Ayan Chakrabarti, Keigo Hirakawa, Todd Zickler
Paper: IEEE Xplore URL: Project website Learning Object Color Models from Multi-view Constraints CVPR 2011
Trevor Owens, Kate Saenko, Ayan Chakrabarti, Ying Xiong, Todd Zickler, Trevor Darrell
Paper: IEEE Xplore Analyzing Spatially-varying Blur CVPR 2010
Ayan Chakrabarti, Todd Zickler, William T. Freeman
Paper: IEEE Xplore URL: Project website An Empirical Camera Model for Internet Color Vision BMVC 2009
Ayan Chakrabarti, Daniel Scharstein, Todd Zickler
Paper: PDF URL: Project website Color Constancy Beyond Bags of Pixels CVPR 2008
Ayan Chakrabarti, Keigo Hirakawa, Todd Zickler
Paper: IEEE Xplore Effective Separation of Sparse and Non-sparse Image Features for Denoising ICASSP 2008
Ayan Chakrabarti, Keigo Hirakawa
Paper: IEEE Xplore Super-Resolution of Face Images Using Kernel PCA-Based Prior IEEE Transactions on Multimedia, June 2007
Ayan Chakrabarti, A.N. Rajagopalan, Rama Chellappa
Paper: IEEE Xplore Visual Inference with Statistical Models for Color and Texture Ph.D. Dissertation, School of Engineering and Applied Sciences, Harvard University, 2011
Acknowledgments: My work is / has been supported generously by the National Science Foundation ( IIS:1618021), and by gifts from Adobe and NVIDIA.
code reference implementations
data Shape From Shading Evaluation Set
Images of uniform-albedo Lambertian objects under directional lighting, with ground truth normals computed using multi-image photometric stereo.
Download Page Camera Color Calibration Datasets
RAW-JPEG pairs of calibration targets captured with different consumer cameras. Useful for the analysis and calibration of camera color processing pipelines.
PAMI 2014 version / BMVC 2009 version (more cameras, but fewer sampled colors per camera) Real-world Hyperspectral Image Database
Database of hyperspectral images of natural scenes, with 31 narrow-band spectral measurements in the visible range at each pixel.
Download Page Motion Blur Database
Database of still images with motion-blurred moving subjects, to qualitatively evaluate algorithms that seek to estimate spatially-varying blur.
Download Page Copyright © 2017 Ayan Chakrabarti.