Akshay Mundra

DENSO ADAS

Biography

I am a Computer Vision Researcher holding a Master of Science in Visual Computing from Saarland University and a Bachelor of Engineering in Electronics and Instrumentation Engineering from BITS Pilani.

My primary interests lie in the areas of Imaging, Computer Vision, and Computer Graphics. For my Master's thesis at MPI Informatics, I focused on exploring the intersection of these fields. Specifically, I developed a method to generate photorealistic 3D hand avatars from 2D images. This research aimed to create a personalized and intuitive interface for VR and AR applications.

I am excited about leveraging my expertise to explore new frontiers in Visual Computing.


Computer Vision Engineer DENSO ADAS 2023 - Present	Graduate Researcher Max-Planck-Institut für Informatik 2021 - 2023	Computer Vision Engineer DreamVu Inc. 2018-2020	Undergraduate Researcher Computer Vision Center, Barcelona Fall 2017

Publications

LiveHand: Real-time and Photorealistic Neural Hand Rendering

Akshay Mundra, Mallikarjun B R, Jiayi Wang, Marc Habermann, Christian Theobalt, Mohamed Elgharib

ICCV 2023

Abstract ▾

The human hand is the main medium through which we interact with our surroundings. Hence, its digitization is of uttermost importance, with direct applications in VR/AR, gaming, and media production amongst other areas. While there are several works modeling the geometry of hands, little attention has been paid to capturing photo-realistic appearance. Moreover, for applications in extended reality and gaming, real-time rendering is critical. We present the first neural- implicit approach to photo-realistically render hands in real-time. This is a challenging problem as hands are textured and undergo strong articulations with pose-dependent effects. However, we show that this aim is achievable through our carefully designed method. This includes training on a low- resolution rendering of a neural radiance field, together with a 3D-consistent super-resolution module and mesh-guided sampling and space canonicalization. We demonstrate a novel application of perceptual loss on the image space, which is critical for learning details accurately. We also show a live demo where we photo-realistically render the human hand in real-time for the first time, while also modeling pose- and view-dependent appearance effects. We ablate all our design choices and show that they optimize for rendering speed and quality.

Code Project Page arXiv Thesis

Projects

Imitation Learning for Autonomous Driving Thesis Code

Developed an end-to-end perception and navigation system for an autonomous vehicle in a virtual environment using imitation learning.
Fine-tuned the model with real-world data to bridge the sim-to-real domain gap.

Ray tracer in C++ Code Rendering Competition

Built a ray tracer in C++ with salient features such as acceleration structures and distribution ray tracing.
Showcased the ray tracer in a rendering competition.

Automated Traffic Control Monitoring Report Code

Developed an end-to-end ML pipeline to estimate traffic density from smartphone and vehicle dashcam images.
Pre-processed an in-the-wild dataset and trained a lightweight MobileNetV2 model on it.
Deployed the model on an android compatible application for real-time inferencing.

Multi-Frame Super Resolution for smartphone photography Report Code

Generated bursts of low-resolution images synthetically, imitating tremors common in hand-held photography.
Applied transfer learning to a Residual Feature Attention based model for multi-frame super-resolution.

Awards

Smart City Expo World Congress 2017

Showcased the work done on autonomous driving at the Smart City Expo World Congress in Barcelona.

APOGEE Awards

Secured first position in APOGEE 2016 for the project Autonomous Solar Powered Grass Cutter. Received the prestigious APOGEE Award from BITSAA International for outstanding performance.

MCN Scholarship at BITS Pilani

Received partial sponsership for tuition fees for all four years of undergraduate study at BITS Pilani.