Shreyas Sachan
M.Sc. in Visual Computing
I am a Master of Science student in Visual Computing at Universität des Saarlandes, currently writing my master's thesis on "3D Mesh Reconstruction from a Single Event Camera." I have the privilege of conducting this research within the 4D and Quantum Computer Vision Group at the Visual Computing and Artificial Intelligence Department of Max Planck Institute for Informatics, under the expert guidance of Prof. Dr. Christian Theobalt and Dr. Vladislav Golyanik.
Previously, I was a Research Assistant at MPI-INF under Dr. Mohamed Elgharib, focusing on 3D face and hand capture, and a Project Scientist at IIT Delhi under Prof. Srikanta Bedathur. Prior to that, I completed my Bachelor's thesis on neural network quantization at IIT Delhi under Prof. Kolin Paul, following my undergraduate studies at the Cluster Innovation Centre, University of Delhi.
Interests & Hobbies: In addition to my academic interests, I enjoy playing games, mainly FPS or MOBA games. In my free time, I also enjoy cooking, travelling and reading books.
Research Interests
- 3D Reconstruction & Neural Rendering
- Event-based (Neuromorphic) Vision
- Deep Learning for 3D Computer Vision
Research Experience
Master's Thesis: 3D Mesh Reconstruction from a Single Event Camera
Max Planck Institute for Informatics
Developed EventNeuS, the first neural framework to extract high-quality 3D meshes solely from monocular event streams. This work introduced SDF-based representations for event data and utilized spherical harmonics to improve reconstruction quality, outperforming the previous state-of-the-art method with a 34% lower Chamfer distance.
Research Assistant
Max Planck Institute for Informatics (MPI-INF)
Designed and deployed a multi-view camera rig for high-precision 3D face and hand gesture capture. Implemented camera calibration pipelines and developed algorithms for synchronizing and processing multi-view visual data.
Projects
3D Ray Tracer
Implemented a 3D ray-tracer in C++ with Monte Carlo path tracing, realistic lighting, shadows, and multi-threading for photorealistic rendering.
Neural Style Transfer
Implemented Gatys' neural style transfer algorithm using VGG-19 to apply the artistic style of Madhubani paintings to images.
Visual Perception for Visually Impaired
Developed a 3D perception pipeline using RGBD data that translates spatial information into audio and haptic feedback. Won "Best Algorithm" at e-Yantra Ideas Competition, IIT Bombay.
Selected Publications
EventNeuS: 3D Mesh Reconstruction from a Single Event Camera
Shreyas Sachan, Viktor Rudnev, Mohamed Elgharib, Christian Theobalt, and Vladislav Golyanik. In International Conference on 3D Vision (3DV), 2026.
An Intelligent Model Based on Integrated Inverse Document Frequency and Multinomial Naive Bayes for Current Affairs News Categorisation
Kumar S., Sharma A., Reddy B.K., Sachan S., Jain V. & Singh J. (2021). International Journal of System Assurance Engineering and Management.
[DOI]