I am a researcher at the Computer Vision Lab, ETH Zurich, working with Prof. Luc Van Gool. My research interests include Computer Vision, Visual-SLAM, Unsupervised Learning, and Optimization Methods. I received my Doctoral degree (PhD) in 2015, and a Master's degree in computer vision in 2012, form University of Bourgogne, France. I also worked as a research scholar at University of Strasbourg, France, form 2013 to 2015, while devising global and local methods for 2D-3D registration problems.
Given correspondences between two instances of the same scene, registration parameters are searched such that the consensus among them is maximized.
Camera intrinsics are obtained without using any known pattern in the scene. Autocalibration only uses multiple images taken by a moving camera. While devising globally optimal methods, the practical issues of noise/outliers are also considered.
Static scene parts are reconstructed and semantically statemented directly in the 3D space. Similarly, 3D trajectories of dynamic objects are segmented using their motion behaviours, to which semantic labels are assigned using 2D-to-3D label transfer.
Image-based localization has an advantage of being inexpensive and efficient. This requires to maintain a large amount of reference images. Reference images are summarized as landmarks, such that image-based navigation can be performed.
Stable training of image/video generative adversarial networks (GANs) is difficult. Improvements in this regard are made by designing better network architectures and loss functions. Improved models are used for multi-functional applications.
More powerful neural networks capable of multi-tasking are very desirable. On the other hand, networks aware of the quality of their prediction are also in very high demand, especially for sensitise tasks such as autonomous driving, and biometrics.
Projects & Seminars: Visual Localization.
Switzerland Master’s level labs and tutorials in Computer Vision.
Centre Universitaire CondorcetMaster’s level lectures in Computer Vision, 3D Vision, C++ programming, and Digital signal processing.
Co-supervisor of more than Twenty Master’s theses, more than Twenty Master’s semester projects, and four research internships.
Area Co-chair for Multimedia and Vision track, IEEE International Conference on Multimedia Information Processing and Retrieval (MIPR) 2018.
Reviewer for various editions of Vision and Robotics related Journal and Conferences: IJCV, CVPR, ICRA, IROS, ICCV, ECCV, PR, ICIP, ICPR, ITSM, and IET-CV.
Organizing committee member of the AI winter schools in Nepal, 2018 and 2019.
Co-founder of a non-profit organization, Nepal Applied Mathematics and Informatics Institute for Research.
VarCity was an ERC research project at the Computer Vision Lab, ETH Zurich. Its goal was to obtain 3D reconstructions using only images, together with the understanding of city static composition, dynamic movement, and virtual life on social media.Read More
The main objective of ENCORE is to increase the share of renovated stock by providing effective and affordable BIM for renovation works. The Computer Vision Lab is involve in providing the required BIM model using only the visual data.Read More
NAAMII is a profit-not-distributing research organization which plays a pivotal role in building a strong foundation of research and support the ecosystem of higher education and innovation in the field of informatics, applied mathematics and AI.Read More
The best paper award by IEEE Computer Society: Domain Agnostic Feature Learning for Image and Video Based Face Anti-spoofing, CVPRW 2020.
A most interesting publication of the week by DeepAI: Sliced Wasserstein Generative Models(CVPR 2019). J. Wu, Z. Huang, D. Acharya, W. Li, J. Thoma, D. P. Paudel, and L. Van Gool. Awarded NVIDIA GPU grant for 2016–2017.
Invited by the International Journal of Computer Vision for Special Issue, Best of ICCV 2015:Robust and optimal registration of image sets and structured scenes via sum-of-squares polynomials(IJCV 2018). D. P. Paudel, A. Habed, C. Demonceaux, and P. Vasseur.
Accepted with Travel Support for ICCV Doctoral Consortium, 2015, Santiago, Chile.
Selected for, and Attended the International Computer Vision Summer School (ICVSS), 2014.
Doctoral Research scholarship awarded by French National Research Agency (2012-2015).Best Master’s Thesis awarded by PAL Robotics on Vibot Day, June 14th, 2012, Girona, Spain.
Master Research scholarship granted by Conseil regional de Bourgogne, France (2011-2012).
National level Competitions during undergraduate studies in India: •Awarded three first prizes in Electronic circuit design and Programming (2007 – 2008). •Awarded two second prizes in Robotics and Technical quiz (2007 – 2008).
Nepal Aid Scholarship to study Electronics and Communication Engineering granted by the Government of India (rank: 3 out of≈2,000 in Nepal).