Hi, Welcome to my homepage!!!
I am a Research Scientist at NVIDIA Research and part of the Machine Learning and Perception group headed by Jan Kautz. Prior to that, I completed my PhD in Computer Science (2014-2018) from the University of Bonn, Germany, under the supervision of Prof. Juergen Gall. I completed my M.Sc. degree in Information Technology (2011-2013) with majors in Image and Video Signal Processing from Tampere University of Technology (TUT), Finland. Alongside, I also worked at Nokia Research Center as an External Computer Vision Researcher. I completed my B.Sc. degree in Computer Engineering from COMSATS Institute of Information Technology, Pakistan, and later also worked there as a Research Associate at the Computer Vision Research Group during 2010-2011.
Research Interests: Computer Vision, Deep Learning, human pose estimation, object detection and recognition, action recognition, face recognition, person re-identification, multi-target tracking
I am looking for motivated students for internships at NVIDIA Research. Experience with 2D/3D human body pose estimation is a plus. If interested drop me an email.
Two papers accepted to CVPR'20. One on weakly-supervised 3D human pose estimation and the other on self-supervised viewpoint learning. More info to follow.
A paper on few-show adaptive gaze estimation has been accepted to ICCV'19. Code coming soon!
My dissertation has won the DAGM MVTec Dissertation Award 2019. DAGM (German Association for Pattern Recognition) represents the German community for pattern recognition including computer vision and machine learning.
I have joined Jan Kautz's lab at Nvidia Research as a post-doctoral Research Scientist.
Successfully defended my PhD with the highest possible distinction (summa cum laude).
Received the Best Poster Award for our work on hand pose estimation at the 4th International Workshop on Observing and Understanding Hands in Actions (HANDS), 2018
Our work on 2D and 3D hand pose estimation is accepted to ECCV 2018.
JointFlow paper on multi-person pose tracking is accepted to BMVC-2018.
JointFlow paper on multi-person pose tracking is available online.
A paper on 2D and 3D hand pose estimation from RGB images is available on ArXiv.
We are organizing 2nd PoseTrack Challenge and Workshop in conjunction with ECCV-2018 in Munich. Stay tuned at http://www.posetrack.net for more info!!
I will spend next six months as Research Intern in Jan Kautz's lab at Nvidia Research.
Evaluation server for PoseTrack challenge is now open. All the best to all participants.
Pre-release of PoseTrack dataset is available now. Grab your copy here.
Webpage for PoseTrack Challenge is down as we are moving our website to a new server. Apologies for the downtime. Will be accessible again soon.
The extended version of our CVPR-2016 paper on 3D human pose estimation is available. [arXiv].
Webpage for PoseTrack Challenge is up!
We are organizing the PoseTrack Challenge on Human Pose Estimation and Tracking in the Wild at ICCV-2017. Stay tuned at http://www.posetrack.net for more info!!
Our paper on Multi-Person Pose Estimation and Tracking accepted to CVPR-2017, arXiv. Source code, models, and dataset coming Soon...
Our paper on action conditioned human pose estimation accepted to FG-2017, arXiv. Project Page (Coming Soon).
A paper on multi-person pose tracking is available on arXiv.
A paper on multi-person human pose estimation accepted to ECCV, Workshop on Crowd Understanding, 2016. Project Page.
A paper on action conditioned human pose estimation is available on arXiv.
A paper on 3D human pose estimation accepted to CVPR 2016!
Project page is up.
Starting doctoral studies at the Computer Vision Research Group (Prof. Dr. Juergen Gall), University of Bonn, Germany