Arihant Gaur

I am a fourth-semester student at Saarland University (M.Sc. Computer Science), currently working on my thesis at 4DQV lab under Dr. Vladislav Golyanik

This semester, I am tutoring two courses: 3D and 4D Computer Vision and Image Acquisition Methods.

Email  /  CV  /  LinkedIn  /  Google Scholar  /  X (Twitter)  /  GitHub

profile photo
Areas of Interest

My research experience has mostly been on 3D computer vision. At Metacube, I worked on 3D skeleton based action recognition. Previously at MERL, I've worked on neural field representations for 3D objects. Please refer to my CV for more details, or mail for questions.

Research
origin Oriented-grid Encoder for 3D Implicit Representations
Arihant Gaur, Gonçalo Dias Pais, Pedro Miraldo
3DV, 2024
arxiv / patent (US) / patent (JP)

Introduces explicit 3D grid based encoding for structured 3D implicit representation
origin Open-Set Multi-Source Multi-Target Domain Adaptation
Rohit Lal, Arihant Gaur, Aadhithya Iyer, Muhammed Abdullah Shaikh, Ritik Agrawal, Shital Chiddarwar
NeurIPS, 2021 (Pre-registration workshop)
arxiv / code

Proposed concept for Multi-Source Multi-Target (nSmT) domain adaptation
origin Health Estimation of Electrical Transformer
Arihant Gaur, Aadhithya Iyer, Ganesh Rohit, Nishit Khandelwal, Vinit Kavalekar, Vyankatesh Muley
B.Tech. thesis, 2022
PDF

Exploring machine learning techniques to summarize overall health of transformer using various online and offline electrical transformer measurements
Other Projects
project1 NeoTracer: Volumetric path tracer with MIS
October 2024 — February 2025
code / project page

Built ray tracer as a part of Computer Graphics - I (Link) course offered at UdS. Bagged third prize for rendering competition hosted during the course
project2 Basic Structure from Motion
July 2020 — September 2020
code

Undertook small project to implement basic blocks of incremental structure from motion. Initial intent was to re-create COLMAP in python, however soon realized issues with the approach :)
project3 Basic Monocular Visual Odometry
March 2020 — May 2020
code

Again a self-learning project, to understand visual odometry before moving to SfM. Implemented 2D - 2D and 3D - 2D visual odometry
News

  • October 2025: I am tutoring two courses at Saarland University this semester: "3D and 4D Computer Vision" and "Image Acquisition Methods"
  • April 2025: I will start my thesis project at the Max Planck Institute for Informatics (4DQV) this semester
  • February 2025: Website for the submission to rendering competition is live. Check it out here: Link
  • April 2024: I've joined as a masters student at Saarland University.
  • October 2023: I started as a short-time intern at Metacube Software Pvt. Ltd.
  • October 2023: My work at MERL, has been accepted to 3DV 2024.
  • August 2022: Started as a research intern at MERL.
  • May 2022: Finished my Bachelor's degree in Electrical and Electronics Engineering.
  • January 2022: Received Academic Excellence Prize for exhibiting the best performance in the 3rd year of B.Tech. Electrical and Electronics Engg. program.
  • October 2021: My paper "Open-Set Multi-Source Multi-Target Domain Adaptation" got accepted at the NeurIPS'21 pre - registration workshop. Paper: Link, Video: Link
  • May 2021: Started working as a research intern at IST, Lisbon.


    • The website template is the courtesy of Jon Barron | Last Updated: 15-10-2025