Hi, I'm Siddhant! I'm a Postdoctoral Research Fellow working with Prof. Tobias Ritschel and Prof. Anthony Steed from the Virtual Environments and Computer Graphics (VECG) group at University College London (UCL) on the Pipelines project. My research deals with photorealistic rendering for mixed-reality (AR/MR) on head-mounted displays (HMDs) based on human perception, often with the help of advanced machine learning and neural rendering techniques.
I completed my Ph.D. under the supervision of Dr. George Drettakis with the GraphDeco group at INRIA Sophia Antipolis-Méditerranée. In my dissertation I studied photorealistic rendering of real environments using captured media such as digital images and videos. My Ph.D. was supported by the ERC Advanced Grant FUNGRAPH.
I graduated with a Master’s degree in Computer Science from Arizona State University, co-advised by Prof. Robert LiKamWa and Prof. Yezhou Yang. During my master's, I looked into problems at the intersection of visual computing and mobile systems, designing novel systems and algorithms to improve mobile mixed-reality experiences. Previously, I obtained my B.Tech (Hons.) degree in Computer Science and Engineering from IIIT Hyderabad, where I was advised by Prof. P J Narayanan as an honors student. In my thesis, I explored the area of computational photography and digital image representation.
My interest in visual computing stems from my love for photography. I enjoy capturing candid moments and particularly appreciate travel and event photography. Checkout few of my captures on my Photo Blog.
Feel free to reach out if you are interested in my research, want to collaborate, or chat about my photography, travel tales, "the boy who lived" or events from a galaxy far, far away!
Publications
Using appearance to efficiently render and edit captured scenes
Deep Scene-Scale Material Estimation from Multi-view Indoor Captures
Hybrid Image-based Rendering for Free-view Synthesis
Proceedings of the ACM on Computer Graphics and Interactive Techniques. ACM I3D, 2021
Paper Supplemental Video Project Page Code Author Ver. (High Res.) I3D Presentation BibTex
Glossy Probe Reprojection for Interactive Global Illumination
Image-Based Rendering of Cars using Semantic Labels and Approximate Reflection Flow
Proceedings of the ACM on Computer Graphics and Interactive Techniques. ACM I3D, 2020
GLEAM: An Illumination Estimation Framework for Real-time Photorealistic Augmented Reality on Mobile Devices
17th ACM International Conference on Mobile Systems, Applications, and Services. ACM MobiSys, 2019
Oral and Demo presentation. (Best Demo Runner-up Award)
Generating Light Estimation for Mixed-reality Devices through Collaborative Visual Sensing
Real-time Illumination Estimation Using Collaborative Photorealistic Rendering for Mobile Augmented Reality
Poster presentation. (Best Poster Award)
Experience
Research Engineer, GraphDeco INRIA Sophia-Antipolis
February 2019 - October 2019
Refactored an Image Based Rendering software along with reference algorithms for open-source release under the supervision of George Drettakis.
Research Aide, Meteor Studio
January 2016 - December 2018
Studied system optimization for enhanced performance of computer vision and graphics on mobile and embedded platforms with Prof. Robert LiKamWa.
Graduate Teaching Assistant, CIDSE ASU
January 2018 - May 2018
Teaching assistant for CSE470 - Computer Graphics under Dr. Dianne Hansford in Spring 2018.
Research Intern, Perceiving Systems MPI-IS
June 2017 - August 2017
Interned under Prof. Michael Black to study real-time dense optical flow estimation on mobile phones using live camera feed using Convolutional Neural Networks.
Here is the most recent version of my CV and Resume (October 2024).
To get a broader idea about my projects and past experience, you can also visit my LinkedIn and Github profiles.
Software
System for Image-Based Rendering (SIBR)
The System for Image-Based Rendering or SIBR, is a specialized collection of libraries and toolkits for quickly implementing Image-Based Rendering (IBR) algorithms, and includes implementations of several published IBR papers, mainly from Inria and UCL, but also (re-implementations) of projects from other research groups. My contributions include designing the current core/projects structure in addition to clean-up towards code release.
Torch to Caffe2 Model Converter
Step-by-Step instructions to convert a Torch model to Caffe2 model. The software takes a pre-trained Torch model and its network parameters as input and creates its copy to run with a Caffe2 network.
Arachnophobia
Arachnophobia is a 2D game using OpenGL. It is one of the first graphics code I wrote for my introductory computer graphics course assignment. You can shoot spiders, or catch them in buckets to save your base! The code gives an overview of all basic OpenGL calls required to create a scene and render it from scratch using primitive geometry.
Contact
Email (best way to reach me): siddhant dot prakash16 at gmail dot com
LinkedIn: Siddhant Prakash LinkedIn