Bio
Oncel Tuzel is a principal researcher and research manager at AI Research group in Apple. Previously he served as senior principal research scientist at Mitsubishi Electric Research Laboratories (MERL), Cambridge. He received his BS and the MS degrees in computer engineering from the Middle East Technical University, Ankara, Turkey in 1999 and 2002 respectively, and the Ph.D. from the computer science department at Rutgers University in 2008. Prior to his Ph.D., he worked as a lead software engineer for four years in Ankara, Turkey developing 3D games and simulations.
His research interests are in machine learning, computer vision and robotics. Previously, he has contributed broadly to object and motion detection, pose estimation, object tracking, image enhancement and super-resolution, facial analysis, and video analytics with commercial applications including robotic assembly systems, video surveillance, medical systems, intelligent transportation, car navigation, satellite systems, automation, and consumer electronics. His current research topics include perception and prediction using deep neural networks, learning from simulated data, deep generative networks, structured learning, and reinforcement learning.
He has co-authored over 50 peer-reviewed publications and holds over 35 US and international patents. His work has received the best paper award in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), the best paper runner-up award in 2007 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), honorable mention award in 2015 Robotics Science and Systems Conference (RSS), and the 2014 R&D 100 award-- awarded to 100 most innovative technology introduced in 2013.
email / Google Scholar / LinkedIn / Twitter / GitHub
News
- Apple’s first deep learning paper authored by my team, "Learning from simulated and unsupervised images through adversarial training", won the best paper award in top AI conference CVPR 2017.
Awards
- Best paper award in leading machine learning and computer vision conference, IEEE Conf. on Comp. Vis. and Pattern Rec. (CVPR), out of 2620 papers, 2017
- Best paper runner-up award in leading machine learning and computer vision conference, IEEE Conf. on Comp. Vis. and Pattern Rec. (CVPR), out of 1300 papers, 2007
- Best paper honorable mention in leading robotics conference, Robotics: Science and Systems (RSS), 2015
- R&D 100 award for MELFA 3D Vision System (Awarded to 100 most innovative technology introduced in 2013), 2014
- MERL research excellence award for contributions to business of intelligent transportation systems, 2013
- Mitsubishi Electric Corporation Information Technology R&D Center (ITC) research excellence award for contributions to product development of vehicle recognition systems, 2012
- MERL directors’ award for contributions to sequential data analysis methods for visual surveillance systems, 2010
- MERL directors’ award for contributions to product development of robotic assembly systems, 2009
Publications
2017
- Learning from simulated and unsupervised images through adversarial training Ashish Shrivastava, Tomas Pfister, Oncel Tuzel, Josh Susskind, Wenda Wang, Russ Webb IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Hawaii, 2017, Best Paper Award
- Attentional Network for Visual Object Detection Kota Hara, Ming-Yu Liu, Oncel Tuzel, Amir-massoud Farahmand arXiv, 2017
2016
- Coupled Generative Adversarial Networks Ming-Yu Liu, Oncel Tuzel Neural Information Processing Systems (NIPS), Barcelona, 2016
- Robust Face Alignment Using a Mixture of Invariant Experts Oncel Tuzel , Tim Marks, Salil Tambe European Conf. on Computer Vision (ECCV), Amsterdam, 2016
- Gaussian Conditional Random Field Network for Semantic Segmentation Raviteja Vemulapalli, Oncel Tuzel, Ming-Yu Liu, Rama Chellappa IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Las Vegas, 2016
- Deep Gaussian Conditional Random Field Network: A Model-based Deep Network for Discriminative Denoising Raviteja Vemulapalli, Oncel Tuzel, Ming-Yu Liu IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Las Vegas, 2016
- A Multi-Stream Bi-Directional Recurrent Neural Network for Fine-Grained Action Detection Bharat Singh, Mike Jones, Tim Marks, Oncel Tuzel IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Las Vegas, 2016
- Unsupervised Deep Network Pretraining via Human Design Ming-Yu Liu, Arun Malya, Oncel Tuzel, Stephen Chen IEEE Winter Conf. on Applications of Computer Vision (WACV), 2016
- R-CNN for Small Object Detection Chenyi Chen, Ming-Yu Liu, Oncel Tuzel, Jianxiong Xiao Asian Conference on Computer Vision (ACCV), Taipei, 2016
- Designing a Boosted Classifier on Riemannian Manifolds Fatih Porikli, Oncel Tuzel, and Peter Meer Riemannian Computing in Computer Vision, 2016
- Global-Local Face Upsampling Network Oncel Tuzel , Yuichi Taguchi, John Hershey arXiv, 2016
- High-accuracy user identification using EEG biometrics Toshiaki Koike-Akino, Ruhi Mahajan, Tim K Marks, Ye Wang, Shinji Watanabe, Oncel Tuzel , Philip Orlik IEEE International Conference of the Engineering in Medicine and Biology Society (EMBC), 2016
2015
- Layered Interpretation of Street View Images Ming-Yu Liu, Shuoxin Lin, Srikumar Ramalingam, Oncel Tuzel Robotics: Science and Systems Conference (RSS), 2015, Best Paper Honorable Mention
- Deep Hierarchical Parsing for Semantic Segmentation Abhishek Sharma, Oncel Tuzel, David Jacobs IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Boston, 2015
- Efficient upsampling of natural images Chinmay Hegde, Oncel Tuzel, Fatih Porikli arXiv, 2015
- Robust Face Alignment Using a Mixture of Invariant Experts Oncel Tuzel , Salil Tambe, Tim Marks arXiv, 2015
2014
- Recursive Context Propagation Network for Semantic Scene Labelling Abhishek Sharma, Oncel Tuzel, Ming-Yu Liu Neural Information Processing Systems (NIPS), 2014
- Learning to rank 3D features Oncel Tuzel , Yuichi Taguchi, Ming-Yu Liu, Arvind Raghunathan European Conf. on Computer Vision (ECCV), 2014
- Semi-Supervised Kernel Mean Shift Clustering Saket Anand, Sushil Mittal, Oncel Tuzel, and Peter Meer IEEE Trans. Pattern Anal. Machine Intelligence (PAMI), Vol. 36, No. 1, pp. 99-112, 2014
- Entropy-Rate Clustering: Cluster Analysis via Maximizing a Submodular Function subject to a Matroid Constraint Ming-Yu Liu, Oncel Tuzel, Srikumar Ramalingam and Rama Chellappa IEEE Trans. Pattern Anal. Machine Intelligence (PAMI), Vol. 36, No. 1, pp. 99-112, 2014
- Detecting 3D Geometric Boundaries of Indoor Scenes Under Varying Lighting Jie Ni, Tim Marks, Oncel Tuzel, Fatih Porikli IEEE Winter Conf. on Applications of Computer Vision (WACV), 2014
2013
- Joint Geodesic Upsampling of Depth Images Ming-Yu Liu, Oncel Tuzel, Yuichi Taguchi IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2013
2012
- Motion-Aware Structured Light Using Spatio-Temporal Decodable Patterns Yuichi Taguchi, Amit Agrawal, Oncel Tuzel European Conf. on Computer Vision (ECCV), 2012
- Voting-based Pose Estimation for Robotic Assembly Using a 3D Sensor Changhyun Choi, Yuichi Taguchi, Oncel Tuzel, Ming-Yu Liu, Srikumar Ramalingam IEEE International Conference on Robotics and Automation (IRCA), pp. 1724-1731, May 2012
- Fast Object Localization and Pose Estimation in Heavy Clutter for Robotic Assembly Ming-Yu Liu, Oncel Tuzel, Ashok Veeraraghavan, Rama Chellappa, Yuichi Taguchi, Tim Marks International Journal of Robotics Research (IJRR), May 2012
- Variable Focus Video: Reconstructing Depth and Video for Dynamic Scenes Nitesh Shroff, Ashok Veeraraghavan, Yuichi Taguchi, Oncel Tuzel, Amit Agrawal Rama Chellappa IEEE International Conference on Computational Photography (ICCP), Pages 1-9, April 2012
2011
- Compressed Inference for Probabilistic Sequential Models Gungor Polatkan and Oncel Tuzel Uncertainty in Artificial Intelligence(UAI), Barcelona, Spain, 2011
- Entropy Rate Superpixel Segmentation Ming-Yu Liu, Oncel Tuzel, Srikumar Ramalingam, Rama Chellappa IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Colorado Springs, 2011
- Finding a Needle in a Specular Haystack Nitesh Shroff, Yuichi Taguchi, Oncel Tuzel, Ashok Veeraraghavan, Srikumar Ramalingam and HaruhisA Okuda IEEE International Conference on Robotics and Automation (ICRA), Shanghai, China, 2011
2010
- P2Pi: A Minimal Solution for Registration of 3D Points to 3D Planes Srikumar Ramalingam, Yuichi Taguchi, Tim Marks and Oncel Tuzel European Conf. on Computer Vision (ECCV), Crete, Greece, 2010
- Image invariants for smooth reflective surfaces Aswin Sankaranarayanan, Ashok Veeraraghavan, Oncel Tuzel and Agrawal. European Conf. on Computer Vision (ECCV), Crete, Greece, 2010
- Fast directional chamfer matching Ming-Yu Liu, Oncel Tuzel, Ashok Veeraraghavan and Rama Chellappa IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), San Francisco, 2010
- Specular surface reconstruction from sparse reflection correspondences Aswin Sankaranarayanan, Ashok Veeraraghavan, Oncel Tuzel and Agrawal IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), San Francisco, 2010
- Pose estimation in heavy clutter using multi flash camera Ming-Yu Liu, Oncel Tuzel, Ashok Veeraraghavan and Rama Chellappa, Amit Agrawal and Haruhisa Okuda IEEE International Conference on Robotics and Automation (ICRA), Anchorage, Alaska, 2010
2009
- Kernel methods for weakly supervised mean shift clustering Oncel Tuzel , Fatih Porikli and Peter Meer IEEE Intl. Conf. on Computer Vision (ICCV), Kyoto, Japan, 2009 (oral presentation)
- A caGrid-enabled, learning based image segmentation method for histopathology specimens David Foran, Lin Yang, Oncel Tuzel, Wenjin Chen, Jun Hu, Tahsin Kurc, Renato Ferreira, Joel Saltz IEEE International Symposium on Biomedical Imaging (ISBI), Boston, 2009
- PathMiner: A web-based telemicroscopy, intelligent archiving and image analysis system Lin Yang, Oncel Tuzel, Peter Meer, Wenjin Chen, Gratan Salaru, Goodell, L. and David Foran IEEE Transaction on Information Technology in Biomedicine. Vol. 13, No. 3, pp. 291-299, 2009
2008
- Automatic Image Analysis of Histopathology Specimens Using Concave Vertex Graph Lin Yang, Oncel Tuzel, Peter Meer and David Foran International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), New York, U. S. A., 2008
- Learning on Lie groups for invariant detection and tracking Oncel Tuzel , Fatih Porikli and Peter Meer IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Anchorage, Alaska, 2008
- Pedestrian detection via classification on Riemannian manifolds Oncel Tuzel , Fatih Porikli and Peter Meer IEEE Trans. Pattern Anal. Machine Intelligence (PAMI), Vol. 30, pp. 1713-1727, 2008
2007
- Human detection via classification on Riemannian manifolds Oncel Tuzel , Fatih Porikli and Peter Meer IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Minneapolis, MN, 2007 (oral presentation), Best Paper Award Runner-up
- Classification of hematologic malignancies using texton signatures Oncel Tuzel , Lin Yang, Peter Meer and David Foran IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Anchorage, Alaska, 2008
2006
- Covariance tracking using model update based on Lie algebra Fatih Porikli, Oncel Tuzel, Peter Meer IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), New York, NY, 2006
- Region covariance: A fast descriptor for detection and classification Oncel Tuzel , Fatih Porikli, Peter Meer European Conf. on Computer Vision (ECCV), Graz, Austria, Vol. 2, 2006, pp. 589–600
- Fast construction of covariance matrices for arbitrary size image windows Oncel Tuzel , Fatih Porikli IEEE International Conference on Image Processing (ICIP), pp.1581-1584, 2006
2005 & earlier
- Simultaneous multiple 3D motion estimation via mode finding on Lie groups Oncel Tuzel , Raghav Subbarao, Peter Meer IEEE Intl. Conf. on Computer Vision (ICCV), Beijing, China, Vol.1 pp. 18–25, 2005 (oral presentation)
- Multi-kernel object tracking Fatih Porikli, Oncel Tuzel IEEE Int. Conf. on Multimedia and Expo (ICME), Amsterdam, Netherlands, pp. 1234-1237, 2005
- Bayesian background modeling for foreground detection Oncel Tuzel , Fatih Porikli ACM international workshop on Video surveillance & sensor networks, Singapore, pp. 55–58, 2005
- Object tracking in low-frame-rate video Oncel Tuzel , Fatih Porikli SPIE Image and Video Communications and Processing, Vol. 5685, pp. 72-79, 2005
- A Bayesian approach to background modeling Oncel Tuzel , Fatih Porikli, Peter Meer IEEE Int. Workshop on Machine Vision for Intelligent Vehicles, San Diego, CA, 2005
- Vision-based single-stroke character recognition for wearable computing Omer F. Ozer, Oguz Ozun, Oncel Tuzel, Volkan Atalay and Enis Cetin IEEE Intelligent Systems, Vol.16 Issue 3 pp. 33-37, 2001
Patents
- Image Upsampling using Global and Local Constraints, Oncel Tuzel, Yuichi Taguchi, pending 2016
- System and Method for Semantic Segmentation using Gaussian Random Field Network, Raviteja Vemulapalli, Oncel Tuzel, Ming-Yu Liu, pending 2016
- Method and System for Detecting Actions in Videos, Mike Jones, Tim Marks, Oncel Tuzel, Singh, B., pending 2016
- Method and System for Denoising Images Using Deep Gaussian Conditional Random Field Network, Oncel Tuzel, Ming-Yu Liu, Raviteja Vemulapalli, pending 2015
- System and Method for Testing and Evaluating Vehicle Components, Zafer Sahinoglu, Oncel Tuzel, pending 2015
- Method for Training Classifiers to Detect Objects Represented in Images of Target Environments, Oncel Tuzel, Jay Thornton, pending 2015
- Method for Labeling Images of Street Scenes, Ming-Yu Liu, Ramalingam S., Oncel Tuzel, pending 2015
- Method for Estimating Locations of Facial Landmarks in an Image of a Face using Globally Aligned Regression, Oncel Tuzel, Salil Tambe Tim Marks, pending 2015
- Method for Semantically Labeling an Image of a Scene using Recursive Context Propagation, Oncel Tuzel, Ming-Yu Liu, Abhishek Sharma, pending 2014
- Method for detecting geometric edges of a scene under multiple illumination conditions, Tim Marks, Oncel Tuzel, Fatih Porikli, Jie Ni, 2016, 9,418,434
- Method for factorizing lighting of a scene into basis images, Oncel Tuzel , Tim Marks, Fatih Porikli, Jie Ni, 2016, 9,384,553
- Image Denoising Using a Library of Functions, Oncel Tuzel, Jay Thornton, Van Baar, J., 2016, 9,262,810
- Method for Determining Object Poses Using Weighted Features, Oncel Tuzel, Ming-Yu Liu, Yuichi Taguchi, Arvind Raghunathan, 2016, 9,280,827
- Specular Edge Extraction Using Multi-Flash Imaging, Yuichi Taguchi, Oncel Tuzel, Srikumar Ramalingam, 2016, 8,913,825
- Method for Detecting Objects in Stereo Images, Ming-Yu Liu, Oncel Tuzel, 2015, 9,195,904
- Method for Increasing Resolutions of Depth Images, Oncel Tuzel, Ming-Yu Liu, Yuichi Taguchi, 2015, 8,983,177
- Method and System for Generating Structured Light with Spatio-Temporal Patterns for 3D Scene Reconstruction, Yuichi Taguchi, Amit Agrawal, Oncel Tuzel, 2014, 8,805,057
- Voting-based Pose Estimation for 3D Sensors, Yuichi Taguchi, Oncel Tuzel, Srikumar Ramalingam, Changhyun Choi, Ming-Yu Liu, 2014, 8,908,913
- Upsampling of natural images, Oncel Tuzel, Fatih Porikli, Chinmay Hegde, 2013, 8,620,073
- Appearance and Context Based Object Classification in Images, Oncel Tuzel, Pillai, J., 2014, 8,718,362
- Method and System for Generating Structured Light with Spatio-Temporal Patterns for 3D Scene Reconstruction, Yuichi Taguchi, Amit Agrawal, Oncel Tuzel, 2014, 8,805,057
- Object detecting with 1D range sensors, Oncel Tuzel, Gungor Polatkan, 2014, 8,824,548
- Method for Segmenting Images Using Superpixels and Entropy Rate Clustering, Oncel Tuzel, Srikumar Ramalingam, Ming-Yu Liu, 2013, 8,428,363
- Object detection in depth images, Mike Jones, Oncel Tuzel, Si, W., 2013, 8,406,470
- Method for Determining Compressed State Sequences, Oncel Tuzel, Gungor Polatkan, 2013, 8,405,531
- Camera and Method for Focus Based Depth Reconstruction of Dynamic Scenes, Ashok Veeraraghavan, Nitesh Shroff, Yuichi Taguchi, Oncel Tuzel, 2013, 8,432,434
- Method for Reconstructing Surfaces of Specular Object from Sparse Reflection Correspondence, Ashok Veeraraghavan, Oncel Tuzel, Amit Agrawal, Aswin Sankaranarayanan, 2012, 8,229,242
- Determining Points of Parabolic Curvature on Surfaces of Specular Objects, Ashok Veeraraghavan, Oncel Tuzel, Amit Agrawal, Aswin Sankaranarayanan, 2012, 8,155,447
- Method and System for Determining Poses of Objects, Oncel Tuzel, Ashok Veeraraghavan, 2012, 8,306,314
- Method and system for determining poses of specular objects, Srikumar Ramalingam, Ashok Veeraraghavan, Yuichi Taguchi Oncel Tuzel, Nitesh Shroff, 2012, 8,165,403
- Method for Clustering Samples with Weakly Supervised Kernel Mean Shift Matrices, Oncel Tuzel, Fatih Porikli, 2012, 8,296,248
- Method and System for Detecting and Tracking Objects in Images, Oncel Tuzel, Fatih Porikli, 2011, 7,961,952
- Detecting Moving Objects in Video by Classifying on Riemannian Manifolds, Fatih Porikli, Oncel Tuzel, 2011, 7,899,253
- Method for Classifying Data Using an Analytic Manifold, Fatih Porikli, Oncel Tuzel, 2010, 7,724,961
- Method for Constructing Covariance Matrices from Data Features, Fatih Porikli, Oncel Tuzel, 2010, 7,720,289
- Method for Tracking Objects in Videos Using Covariance Matrices, Fatih Porikli, Oncel Tuzel, 2009, 7,620,204
- Modeling Low Frame Rate Videos with Bayesian Estimation, Fatih Porikli, Oncel Tuzel, 2008, 7,466,842
- Tracking Objects in Low Frame Rate Videos, Fatih Porikli, Oncel Tuzel, 2008, 7,418,113
- Adaptive Background Image Updating, Dirk Brinkman, Fatih Porikli, Oncel Tuzel, 2007, 7,224,735