Publications

@conference{Xie2020,

title = {Count- and Similarity-aware R-CNN for Pedestrian Detection},

author = {Jin Xie and Hisham Cholakkal and Rao Muhammad Anwer and Fahad Shahbaz Khan and Yanwei Pang and Ling Shao and Mubarak Shah},

url = {https://www.crcv.ucf.edu/wp-content/uploads/2020/07/Publications_Count-and-Similarity-aware-R-CNN-for-Pedestrian-Detection.pdf},

year  = {2020},

date = {2020-08-23},

booktitle = {16th European Conference on Computer Vision},

abstract = {Recent pedestrian detection methods generally rely on additional supervision, such as visible bounding-box annotations, to handle heavy occlusions. We propose an approach that leverages pedestrian count and proposal similarity information within a two-stage pedestrian detection framework. Both pedestrian count and proposal similarity are derived from standard full-body annotations commonly used to train pedestrian detectors. We introduce a count-weighted detection loss function that assigns higher weights to the detection errors occurring at highly overlapping pedestrians. The proposed loss function is utilized at both stages of the two-stage detector. We further introduce a count-andsimilarity branch within the two-stage detection framework, which predicts pedestrian count as well as proposal similarity to identify distinct proposals. Our approach requires neither part information nor visible bounding-box annotations. Experiments are performed on the CityPersons and CrowdHuman datasets. Our method sets a new state-of-the-art on both datasets. Further, it achieves an absolute gain of 2.4% over the current state-of-the-art, in terms of log-average miss rate, on the heavily occluded (HO) set of CityPersons test set, without using additional visible bounding-box supervision. Finally, we demonstrate the applicability of our approach for the problem of human instance segmentation. Code and models are available at: https://github.com/Leotju/CaSe.},

keywords = {Detection, ECCV},

pubstate = {published},

tppubtype = {conference}

}

@conference{Xie2020b,

title = {Count- and Similarity-aware R-CNN for Pedestrian Detection},

author = {Jin Xie and Hisham Cholakkal and Rao Muhammad Anwer and Fahad Shahbaz Khan and Yanwei Pang and Ling Shao and Mubarak Shah},

url = {https://www.crcv.ucf.edu/wp-content/uploads/2020/07/Publications_Count-and-Similarity-aware-R-CNN-for-Pedestrian-Detection.pdf},

year  = {2020},

date = {2020-08-23},

booktitle = {16th European Conference on Computer Vision},

abstract = {Recent pedestrian detection methods generally rely on additional supervision, such as visible bounding-box annotations, to handle heavy occlusions. We propose an approach that leverages pedestrian count and proposal similarity information within a two-stage pedestrian detection framework. Both pedestrian count and proposal similarity are derived from standard full-body annotations commonly used to train pedestrian detectors. We introduce a count-weighted detection loss function that assigns higher weights to the detection errors occurring at highly overlapping pedestrians. The proposed loss function is utilized at both stages of the two-stage detector. We further introduce a count-andsimilarity branch within the two-stage detection framework, which predicts pedestrian count as well as proposal similarity to identify distinct proposals. Our approach requires neither part information nor visible bounding-box annotations. Experiments are performed on the CityPersons and CrowdHuman datasets. Our method sets a new state-of-the-art on both datasets. Further, it achieves an absolute gain of 2.4% over the current state-of-the-art, in terms of log-average miss rate, on the heavily occluded (HO) set of CityPersons test set, without using additional visible bounding-box supervision. Finally, we demonstrate the applicability of our approach for the problem of human instance segmentation. Code and models are available at: https://github.com/Leotju/CaSe.},

keywords = {Detection, ECCV},

pubstate = {published},

tppubtype = {conference}

}

@conference{Jamal2019,

title = {Deep Face Detector Adaptation Without Negative Transfer or Catastrophic Forgetting},

author = {M. Jamal and H. Li and Boqing Gong},

url = {https://www.crcv.ucf.edu/papers/cvpr2018/deep-face.pdf},

year  = {2018},

date = {2018-04-09},

publisher = {IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2018), Salt Lake City, UT, June 18-22, 2018.},

keywords = {Detection},

pubstate = {published},

tppubtype = {conference}

}

@conference{Jamal2019b,

title = {Deep Face Detector Adaptation Without Negative Transfer or Catastrophic Forgetting},

author = {M. Jamal and H. Li and Boqing Gong},

url = {https://www.crcv.ucf.edu/papers/cvpr2018/deep-face.pdf},

year  = {2018},

date = {2018-04-09},

publisher = {IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2018), Salt Lake City, UT, June 18-22, 2018.},

keywords = {Detection},

pubstate = {published},

tppubtype = {conference}

}

@conference{LaLonde2018,

title = {ClusterNet: Detecting Small Objects in Large Scenes by Exploiting Spatio-Temporal Information},

author = {Rodney LaLonde and Dong Zhang and Mubarak Shah},

url = {https://www.crcv.ucf.edu/papers/cvpr2018/3460Final.pdf 

https://www.crcv.ucf.edu/papers/cvpr2018/3460-suppFinal.pdf},

year  = {2018},

date = {2018-04-08},

urldate = {2018-04-08},

publisher = {IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2018), Salt Lake City, UT, June 18-22, 2018},

keywords = {Detection, Drone Video Analysis, UAV Video Analysis},

pubstate = {published},

tppubtype = {conference}

}

@conference{LaLonde2018b,

title = {ClusterNet: Detecting Small Objects in Large Scenes by Exploiting Spatio-Temporal Information},

author = {Rodney LaLonde and Dong Zhang and Mubarak Shah},

url = {https://www.crcv.ucf.edu/papers/cvpr2018/3460Final.pdf 

https://www.crcv.ucf.edu/papers/cvpr2018/3460-suppFinal.pdf},

year  = {2018},

date = {2018-04-08},

urldate = {2018-04-08},

publisher = {IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2018), Salt Lake City, UT, June 18-22, 2018},

keywords = {Detection, Drone Video Analysis, UAV Video Analysis},

pubstate = {published},

tppubtype = {conference}

}

@conference{Hou2016,

title = {An End-to-end 3D Convolutional Neural Network for Action Detection and Segmentation in Videos},

author = {Rui Hou and Chen Chen and Mubarak Shah},

url = {https://www.crcv.ucf.edu/papers/ST-CNN-arXiv.pdf},

year  = {2017},

date = {2017-03-31},

publisher = {Cornell University Library},

keywords = {Detection, Segmentation},

pubstate = {published},

tppubtype = {conference}

}

@conference{Hou2016c,

title = {An End-to-end 3D Convolutional Neural Network for Action Detection and Segmentation in Videos},

author = {Rui Hou and Chen Chen and Mubarak Shah},

url = {https://www.crcv.ucf.edu/papers/ST-CNN-arXiv.pdf},

year  = {2017},

date = {2017-03-31},

publisher = {Cornell University Library},

keywords = {Detection, Segmentation},

pubstate = {published},

tppubtype = {conference}

}

@conference{LaLonde2017,

title = {Fully Convolutional Deep Neural Networks for Persistent Multi-Frame Multi-Object Detection in Wide Area Aerial Videos},

author = {Rodney LaLonde and Dong Zhang and Mubarak Shah},

url = {https://www.crcv.ucf.edu/papers/1704.02694v1.pdf},

year  = {2017},

date = {2017-03-23},

publisher = {Cornell University Library},

keywords = {Detection},

pubstate = {published},

tppubtype = {conference}

}

@conference{LaLonde2017b,

title = {Fully Convolutional Deep Neural Networks for Persistent Multi-Frame Multi-Object Detection in Wide Area Aerial Videos},

author = {Rodney LaLonde and Dong Zhang and Mubarak Shah},

url = {https://www.crcv.ucf.edu/papers/1704.02694v1.pdf},

year  = {2017},

date = {2017-03-23},

publisher = {Cornell University Library},

keywords = {Detection},

pubstate = {published},

tppubtype = {conference}

}

@conference{Hou2017,

title = {Tube Convolutional Neural Network (T-CNN) for Action Detection in Videos},

author = {Rui Hou and Chen Chen and Mubarak Shah},

url = {https://arxiv.org/pdf/1703.10664.pdf},

year  = {2017},

date = {2017-03-22},

publisher = {Cornell University Library},

keywords = {Detection},

pubstate = {published},

tppubtype = {conference}

}

@conference{Hou2017c,

title = {Tube Convolutional Neural Network (T-CNN) for Action Detection in Videos},

author = {Rui Hou and Chen Chen and Mubarak Shah},

url = {https://arxiv.org/pdf/1703.10664.pdf},

year  = {2017},

date = {2017-03-22},

publisher = {Cornell University Library},

keywords = {Detection},

pubstate = {published},

tppubtype = {conference}

}

@conference{Sultani2017,

title = {Automatic Action Annotation in Weakly Labeled Videos},

author = {Waqas Sultani and Mubarak Shah},

url = {https://www.crcv.ucf.edu/papers/CVIU_Manuscript_Waqas_AL.pdf},

year  = {2016},

date = {2016-03-09},

publisher = {Cornell University Library},

keywords = {Classification, Detection},

pubstate = {published},

tppubtype = {conference}

}

@conference{Sultani2017d,

title = {Automatic Action Annotation in Weakly Labeled Videos},

author = {Waqas Sultani and Mubarak Shah},

url = {https://www.crcv.ucf.edu/papers/CVIU_Manuscript_Waqas_AL.pdf},

year  = {2016},

date = {2016-03-09},

publisher = {Cornell University Library},

keywords = {Classification, Detection},

pubstate = {published},

tppubtype = {conference}

}