Chethan Parameshwara

I work as a researcher at Amazon, focusing on Multimodal Generative AI, with a specific emphasis on Diffusion Generative Models and 3D Computer Vision.

Before joining Amazon, I graduated with a Ph.D. in Neuroscience and Artificial Intelligence (NACS) from the University of Maryland, College Park (UMD), where I developed neuroscience-inspired computer vision algorithms for autonomous vehicles.

Email  /  GitHub  /  Google Scholar  /  LinkedIn  /  Twitter

Publications·Experiences·Teaching·Volunteering
Publications

My selected publications are listed here. The complete list of publications can be seen from my Google Scholar page.

DiffPoseNet: Direct Differentiable Camera Pose Estimation
Chethan M. Parameshwara, Gokul Hari, Cornelia Fermüller, Nitin J. Sanket, Yiannis Aloimonos
CVPR, 2022
project page / arXiv / video

We propose a novel differentiable programming technique for camera pose estimation in autonomous driving and flying scenarios.

SpikeMS: Deep Spiking Neural Network for Motion Segmentation
Chethan M. Parameshwara*, Simin Li*, Cornelia Fermüller, Nitin J. Sanket, Matthew S. Evanusa, Yiannis Aloimonos (* equal contribution)
IROS, 2021
project page / code / arXiv / video

We propose a bio-inspired neural network for the motion detection problem using an asynchronous event camera.

NudgeSeg: Zero-Shot Object Segmentation by Repeated Physical Interaction
Chahat Deep Singh*, Nitin J. Sanket*, Chethan M. Parameshwara, Cornelia Fermüller, Yiannis Aloimonos (* equal contribution)
IROS, 2021
project page / arXiv / video

We present a novel approach to segment never-seen objects by physically interacting with them and observing them from different views.

0-MMS: Zero-Shot Multi-Motion Segmentation With A Monocular Event Camera
Chethan M. Parameshwara, Nitin J. Sanket, Chahat Deep Singh, Cornelia Fermüller, Yiannis Aloimonos
ICRA, 2021
project page / code / arXiv / video

We propose a hybrid solution to multi-object motion segmentation using a combination of model-based and deep learning approaches with minimal prior knowledge.

EVPropNet: Detecting Drones By Finding Propellers For Mid-Air Landing And Following
Nitin J. Sanket, Chahat Deep Singh, Chethan M. Parameshwara, Cornelia Fermüller, Guido de Croon, Yiannis Aloimonos
RSS, 2021
project page / code / arXiv / video

We present a deep learning-based solution for detecting propellers (to detect drones) for mid-air landing and following.

EVDodgeNet: Deep Dynamic Obstacle Dodging with Event Cameras
Chethan M. Parameshwara*, Nitin J. Sanket*, Chahat Deep Singh, Ashwin V. Kuruttukulam, Cornelia Fermüller, Davide Scaramuzza, Yiannis Aloimonos (* equal contribution)
ICRA, 2020
project page / code / arXiv / video

We present the first deep learning based solution for dodging multiple dynamic obstacles on a quadrotor with a single event camera and onboard computation.

Event-based Moving Object Detection and Tracking
Anton Mitrokhin, Cornelia Fermüller, Chethan M. Parameshwara, Yiannis Aloimonos
IROS, 2018
project page / code / arXiv / video

We present a novel motion compensation approach for moving object tracking with an asynchronous event camera.

Experiences
PontTuset

Amazon
Aug, 2022 -- Present
Applied Scientist II

Research on Diffusion Generative Models and 3D Computer Vision (NeRF, Gaussian Splatting, Neural Rendering, and SLAM).

PontTuset

University of Maryland, College Park (UMD)
Aug, 2017 -- Aug, 2022
Graduate Research Assistant with Prof. Yiannis Aloimonos and Dr. Cornelia Fermüller

Developed neuromorphic motion segmentation and tracking algorithms for high-speed and challenging lighting scenarios and also developed differentiable optimization layers (a combination of model-based and data-driven) for camera pose estimation to improve robustness and generalization across scenes.

PontTuset

SRI International (formerly Stanford Research Institute)
Summer, 2021
Research Intern with Dr. David Zhang and Michael Piacentino

Developed a neuro-inspired learning approach for few-shot image classification with a faster convergence rate (10x) and consumes low memory (20x) than existing image classification algorithms.

PontTuset

Neurala
Summer, 2019
Research Intern with Dr. Anatoli Gorchet and Dr. Matthew Luciw

Developed a lifelong machine learning approach to improve few-shot learning capabilities for object detection tasks and deployed the proposed approach on Neurala's Brain Builder software.

PontTuset

Robot Training Academy
Fall, 2016
Software Engineering Intern

Developed hand gesture tracking software for human-robot interaction in kitchen environments and performed testing of perception software modules on Rethink Baxter robot.

Teaching
panoroma Teaching Assistant, CMSC733 : Geometric Computer Vision

Teaching Assistant, CMSC426 : Computer Vision

Teaching Assistant, CMSC434 : Human Computer Interaction
Volunteering
gsglogo Reviewer: NeurIPS, ICLR, CVPR, ECCV, RAL, ICRA, IROS, IEEE Sensor Journal

NACS Representative, UMD Graduate Student Government (2020-2021)

Co-Chair, NACS Grant Review Committee (2019-2022)

Staff Member, Neuromorphic Cognition Engineering Workshop (July 2018)

Last updated on Oct 20, 2023. Thanks to Jon Barron for this minimalist website template.