Shubham Dokania

me.jpg

Hi! I’m an upcoming PhD candidate at the RTG 2853 Neuroexplicit Models of Language, Vision, and Action at Saarland University, starting from September 2024. I will be co-supervised by Prof. Philipp Slusallek and Prof. Eddy Ilg.

Previously, I was working as a Senior Researcher at Mercedes-Benz Research & Development India working on multi-modal research for vision, language and additional modalities. I have completed my Masters (MS Research, Computer Science) at IIIT Hyderabad, co-supervised by Prof. C. V. Jawahar at CVIT, Prof. Manmohan Chandraker from UCSD . Before that, I was working as a Research Engineer at Mercedes-Benz Research & Development India, on the MBUX Interior Assistant using Computer Vision and Deep Learning.

I finished my undergraduate from Delhi Technological University (Formerly DCE) in 2017, with a degree in Mathematics and Computing. I have been lucky to receive mentorship by some amazing people including Prof. C.V. Jawahar, Prof. Manmohan Chandraker, Dr. Anbumani Subramanian, and Dr. Ganesh Bagler.

I love to explore different problems, especially in computer vision, and experiment with different domains frequently. Sometimes, it ends up in something useful, other times it’s a good learning experience. I try to share my random thoughts on my blog.

Have a look at my CV for more formal details!

news

Aug 10, 2024 Starting as a PhD candidate at the RTG Neuroexplicit @ University of Saarland in September 2024. See the announcement.
Jul 21, 2024 Our paper on Eye Gaze estimation “GazeHELL” accepted to BMVC 2024.
Jun 1, 2023 Graduated from IIIT Hyderabad with MS (Research) in Computer Science and Specialization in Artifical Intelligence.
May 11, 2023 Joined as Senior Researcher at Mercedes-Benz Research & Development India in the Intelligent Interior group.
Nov 7, 2022 Giving a talk on Driving Datasets: Real and Un-Real at CVC, Universitat Autònoma de Barcelona, hosted by Prof. Dimosthenis Karatzas and meet with Dr. Antonio Manuel Lopez.
Nov 4, 2022 Giving a talk on Driving Datasets: Real and Un-Real at University of Catania, hosted by Prof. Giovanni Maria Farinella and Dr. Antonino Furnari.
Oct 10, 2022 IDD-3D: Indian Driving Dataset for 3D Unstructured Road Scenes got accepted at WACV 2023!
Jul 10, 2022 TRoVE: Transforming Road Scene Datasets into Photorealistic Virtual Environments got accepted at ECCV 2022!
May 22, 2022 Student Volunteer at 3D Vision Summer School, IIIT Hyderabad. Conducted a session on 3D data manipulation using Blender, and Graph Neural Networks for Point Cloud analysis.
Aug 1, 2021 Worked as a Student Volunteer for CVIT Summer School in Computer Vision 2021.
Nov 25, 2020 Started working as a Student Researcher (MS) at CVIT, IIIT Hyderabad, advised by Prof. C. V. Jawahar, Prof. Manmohan Chandraker, and Dr. Anbumani Subramanian.
Aug 10, 2020 I’ll be joining IIIT-Hyderabad as a MS (Research) Student.
Feb 18, 2020 Received the Product Innovation Award at Mercedes-Benz R&D India.
Dec 1, 2019 Paper GrahAM accepted at NeurIPS 2019 Workshop: New in ML!

selected publications

  1. GazeHELL: Gaze Estimation with Hybrid Encoders and Localised Losses with weighing
    Dokania, Shubham, Singh, Vasudev,  and Ahmad, Shuaib
    BMVC 2024 2024
  2. IDD-3D: Dataset for Driving on Unstructured Roads
    Dokania, Shubham, Hafez, A.H. Abdul,  Subramanian, Anbumani and 2 more authors
    WACV 2023 2023
  3. TRoVE: Transforming Road Scene Datasets into Photorealistic Virtual Environments
    Dokania, Shubham, Subramanian, Anbumani,  Chandraker, Manmohan and 1 more author
    ECCV 2022 2022
  4. Graph Representation learning for Audio & Music genre Classification
    Dokania, Shubham,  and Singh, Vasudev
    arXiv preprint arXiv:1910.11117 2019