M.Shahzeb Khan Gul

Research Associate at Computational Imaging and Algorithms,
Department of Audio and Media technologies, Fraunhofer IIS,
muhammad.gul@iis.fraunhofer.de, mshahzebkhangul@gmail.com
Google Scholar | Linkedin



Bio

I am Muhammad Shahzeb Khan, a Ph.D. graduate in Informatics from FAU Erlangen-Nürnberg, where my research focused on the neural network-based processing of light fields. I also hold an MS degree in Computer Science from Istanbul Medipol University, where I specialized in super-resolution of light fields using convolutional neural networks. Additionally, I have a BE degree in Electronics from PAF-KIET in Karachi. With over 5 years of experience in programming and deep learning, I have worked extensively with Python and Matlab. I am proficient in TensorFlow, PyTorch, and Keras for deep learning applications and have experience with various programming libraries such as OpenCV, Numpy, and Pandas. Throughout my career, I have contributed to research projects, developed algorithms, and conducted experiments in prestigious institutions such as Fraunhofer IIS and Istanbul Medipol University. Please explore the page to learn more about my publications and professional experiences.



Research Interests

Computer vision, Deep learning, and Plenoptic imaging.



Work Experience

Oct. 2018 - Present. 

Research Associate
Computation Imaging and Algorithms, Fraunhofer IIS, Erlangen (Germany)
- Light-field processing
- Novel view synthesis
- 3D reconstruction
From October, 2021 started working on DSAI project funded by Bavarian Ministry of Economic Affairs, Regional Development and Energy in the amount of 13 million euros for the period 2020 to 2025.

Dec. 2017 - June. 2018. 

Digital Design and Verification Engineer
YongaTek, Istanbul (Turkey)
- RTL based design for the H.264 video decoder.

Aug. 2016 - June. 2018. 

Research Assistant
Istanbul Medipol University, Istanbul (Turkey)

June. 2015 - July. 2016. 

R&D Engineer
Genesis Solutions Pvt Ltd, Karachi (Pakistan)
- Working on multiple Embedded Platforms e.g. FPGA, ARM, Arduino and Microchip Controllers.
- Developed cross-platform GUI controls using C# code shared between Windows and different embedded targets.

Jan. 2015 - June. 2015. 

Lab Engineer
Pakistan Air-force Karachi Institute of Economics and Technology (PAF-KIET), Karachi (Pakistan)
- Conducting labs of undergraduate courses of Electrical Machines and Automation and Robotics.
- Founder and core member of Kiet Robotics Club (KRC).
- Conducting robotics and embedded workshops under KRC.



Publications
Patent

M.Shahzeb Khan Gul, Michel Baetz, Joachim Keinert,
Encoder and Decoder for Encoding and Decoding Images,
U.S. Patent and Trademark Office, U.S. Patent No: 11,546,614.
Link


Book Chapter

M.Shahzeb Khan Gul, Michel Baetz, Joachim Keinert, Laur Fink, Tobias Jaschke, Siegfried Fößel, Florian Goldmann, Nico Prappacher, Matthias Ziegler,
Light Field Processing For Media Applications,
Immersive Video Technologies.
Link


Journals

M.Shahzeb Khan Gul, M. Umair Mukati, Michel Baetz, Soren Forchhammer, Joachim Keinert,
Attention Mechanism-Based Light-Field View Synthesis,
IEEE Access, 2022.
Paper



M.Shahzeb Khan Gul, Bahadir K. Gunturk,
Spatial and Angular Resolution Enhancement of Light Fields Using Convolutional Neural Networks,
IEEE Transactions on Image Processing ( Volume: 27, Issue: 5, May 2018 ).
Paper | Code


Conferences

M.Shahzeb Khan Gul, Hamid Suleman, Michel Baetz, Joachim Keinert,
Stereo Image Compression Using Recurrent Neural Network With A Convolutional Neural Network-Based Occlusion Detection,
International Conference on Pattern Recognition (ICPR 2022), Montréal, Québec, Canada.
Paper



M.Shahzeb Khan Gul, Hamid Suleman, Michel Baetz, Joachim Keinert,
RNNSC: Recurrent Neural Network-Based Stereo Compression using Image and State Warping,
Data Compression Conference (DCC 2022), Snowbird, UT, USA.
Paper



M.Shahzeb Khan Gul, M. Umair Mukati, Michel Baetz, Soren Forchhammer, Joachim Keinert,
Light-Field View Synthesis Using A Convolutional Block Attention Module,
IEEE International Conference on Image Processing (ICIP 2021), Anchorage, Alaska USA.
Paper



M.Shahzeb Khan Gul, Thorsten Wolf, Michel Baetz, Matthias Ziegler, Joachim Keinert,
A high-resolution high dynamic range light-field dataset with an application to view-synthesis and tone-mapping,
IEEE International Conference on Multimedia and Expo (ICME 2020), London, UK || Virtual.
Paper | Dataset



Dingcheng Yue, M.Shahzeb Khan Gul, Michel Baetz, Joachim Keinert, Rafal Mantiuk,
A benckmark of light-field view-interpolation methods,
IEEE International Conference on Multimedia and Expo (ICME 2020), London, UK || Virtual.
Paper | Blender Dataset | Real Dataset | Project Page



M.Shahzeb Khan Gul, Michel Baetz, Joachim Keinert,
Pixel-Wise Confidences for Stereo Disparities Using Recurrent Neural Networks,
British Machine Vision Conference (BMVC 2019), Cardiff, UK.
Paper