Welcome to Computer Vision Lab

Congratulations! Yi Zhu has successfuly defend his PhD dissertation and joined Amazon AI as a research scientist in June.
NEW PAPER: One paper has been accepted at CVPR 2019 (ORAL)! [link]
NEW PAPERS: Three papers have been accepted at ACCV 2018!
NEW PAPER: One paper has been accepted at ACM SIGSPATIAL 2018 (ORAL)! [pdf]
NEW PAPER: One paper has been accepted at CVPR 2018! [pdf]
NEWSOur work on generating ground level views from satellite imagery is covered by MIT technology review, Internet of Business, GIS Lounge, DeepTech. We also had an interview with This Week in Machine Learning & AI and the Youtube link can be found here.
NEW PAPERS: Two papers have been accepted at ICIP 2018
KEYNOTE: Professor Newsam was invited to give a keynote “Geographic knowledge discovery using ground-level images and videos” on 1st Workshop on GeoAI on ACM SIGSPATIAL 2017. [slides]

Recent Papers

[1]Yi Zhu, Karan Sapra, Fitsum A. Reda, Kevin J. Shih, Shawn Newsam, Andrew Tao, Bryan Catanzaro, Improving Semantic Segmentation via Video Propagation and Label Relaxation, The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2019 [ORAL] [pdf]

[2]Yi Zhu, Xueqing Deng and Shawn Newsam, Fine-grained Land Use classification at the city scale using ground-level images, IEEE Transactions on Multimedia, 2019 [pdf]

[3]Yi Zhu, Zhenzhong Lan, Shawn Newsam and Alexander G Hauptmann, Hidden Two-Stream Convolutional Networks for Action Recognition, Asian Conference on Computer Vision (ACCV) 2018 [pdf]

[4]Yi Zhu and Shawn Newsam, Random Temporal Skipping for Multirate Video Analysis, Asian Conference on Computer Vision (ACCV) 2018 [pdf]

[5]Yi Zhu, Jia Xue and Shawn Newsam, Gated Transfer Network for Transfer Learning, Asian Conference on Computer Vision (ACCV) 2018 [pdf]

[6]Xueqing Deng, Yi Zhu and Shawn Newsam, What Is It Like Down There? Generating Dense Ground-Level Views and Image Features From Overhead Imagery Using Conditional Generative Adversarial Networks, ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM SIGSPATIAL GIS) 2018 [ORAL] [pdf]

[7]Yi Zhu, Yang Long, Yu Guan, Shawn Newsam and Ling Shao, Towards Universal Representation for Unseen Action Recognition, The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2018 [pdf]

[8] Yi Zhu and Shawn Newsam, Large-Scale Human Activity Mapping using Geo-Tagged Videos, ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM SIGSPATIAL GIS) 2017 [pdf]

[9] Xueqing Deng and Shawn Newsam, Quantitative Comparison of Open-Source Data for Fine-Grain Mapping of Land Use, ACM SIGSPATIAL workshop on Urban GIS 2017 [pdf]

[10] Weixun Zhou, Shawn Newsam, Congmin Li and Zhenfeng Shao, Learning Low Dimensional Convolutional Neural Networks for High-Resolution Remote Sensing Image Retrieval, Remote Sensing, 2017 [pdf]

[11] Yi Zhu, Zhenzhong Lan, Shawn Newsam, Alexander G. Hauptmann, Hidden Two-Stream Convolutional Networks for Action Recognition, arXiv 2017, [pdf]


Selected Papers

    • Action Recognition

Deep Local Video Feature for Action Recognition
Zhenzhong Lan, Yi Zhu, Alexander G. Hauptmann and Shawn Newsam
IEEE Conference on Computer Vision and Pattern Recognition (CVPR): Workshop on Open Domain Action Recognition (ODAR), 2017 [pdf]
Efficient Action Detection in Untrimmed Videos via Multi-Task Learning
Yi Zhu and Shawn Newsam
IEEE Winter Conference on Applications of Computer Vision (WACV), 2017 [pdf]

    • Optical Flow

Guided Optical Flow Learning
Yi Zhu, Zhenzhong Lan, Shawn Newsam and Alexander G. Hauptmann
IEEE Conference on Computer Vision and Pattern Recognition (CVPR): Workshop on Workshop on Brave New Motion Representations (BNMR), 2017 [pdf]
DenseNet for Dense Flow
Yi Zhu and Shawn Newsam
IEEE International Conference on Image Processing (ICIP) 2017 [pdf]

    • Land Use Classification Using Ground-Level Image

Land use classification using convolutional neural networks applied to ground-level images
Yi Zhu and Shawn Newsam
ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM SIGSPATIAL GIS), 2015 (Best poster award.)[pdf]
Exploring geotagged images for land-use classification
Daniel Leung and Shawn Newsam
ACM International Conference on Multimedia: Workshop on Geotagging and Its Applications in Multimedia, 2012 [pdf]

    • Crowdsourcing What-Is-Where

Proximate Sensing: Inferring What-Is-Where From Georeferenced Photo Collection
Daniel Leung and Shawn Newsam
IEEE International Conference on Computer Vision and Pattern Recognition, 2010 (oral presentation) [pdf]

    • Geographic Image Retrieval

Geographic image retrieval using local invariant features
Yi Yang and Shawn Newsam
IEEE Transactions on Geoscience and Remote Sensing, 2013 [pdf]

  • High-Resolution Overhead Imagery Classification

Bag-of-visual-words and spatial extensions for land-use classificatio
Yi Yang and Shawn Newsam
ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM SIGSPATIAL GIS), 2010 [pdf]