I am a Computer Science MSc student at ETH Zurich, majoring in Machine Intelligence, with a focus on Computer Vision and Mobile Robotics. I did my
bachelors in Computer Science (w/ minor in Data Science) from BITS Pilani, Goa. I am
currently working as a Computer Vision Intern at Scandit, solving interest point detection and tracking
in the MatrixScan pipeline.
I have had the pleasure of working on the following projects with some amazing researchers and engineers at:
I'm currently looking for full-time Machine Learning opportunities! I am excited about advancing research in foundation models and also their applications in robotics and perception systems.
Feel free to check out my
Resume
and drop me an
e-mail
if you want to chat with me!
P.S. I play tennis, and find freedom in running and cycling.
I also enjoy reading and occasionally dabble in chess.
Rafa remains the GOAT for me even if Djokovic goes on to win 30 grand slams.
Computer Vision Student Researcher | Scandit
Jul '24 - Present
ML in the Barcode Tracking Team - With Menelaos Kanakis and Matthias Bloch.
Replacing traditional keypoint detectors with learned detection and matching methods
in the tracking pipeline of Scandit's MatrixScan product,
which leverages SLAM for AR-based inventory management on resource-constrained
devices. Optimizing training paradigms, and exploring lightweight architectures for
efficient on-device inference along with custom evaluation benchmarks.
Graduate Student Researcher | Robotics and Perception Group, UZH
Feb '24 - Present
Research - Under Prof. Dr. Davide Scaramuzza
Developed a unified CLIP-based representation combining
geometry and semantics for Object Goal Navigation in unseen environments.
Demonstrated that integrating CLIP feature based representation along with frontier-based exploration outperforms recent segmentation-based methods, highlighting its robust generalization and zero-shot capabilities.
Graduate Student Researcher | Computer Vision and Geometry Group, ETH
Feb '24 - Present
Research - Under Prof. Dr. Marc Pollefeys
Developed POLD2, a deep learning-based pipeline that jointly detects and describes
both point and line features in images, optimizing feature extraction for 3D vision
tasks like SLAM and pose estimation. By sharing computations between points and
lines, POLD2 achieves a significant 9.5x speedup in inference time compared to
traditional methods, while maintaining comparable accuracy.
Student Researcher | Google Research, India
Aug '22 - Jun '23
Undergraduate Thesis - Accepted to CVPRW '24
Developed a versatile neural network compression toolbox that optimizes for the model's FLOPs via a novel
latency surrogate across a family of compression methods, including pruning and low-rank factorization.
Additionally, optimized on-device latency of large vision models used for OCR tasks in Google Lens, and QR-code
scanning in GooglePay.
Summer Research Intern | Karlsruhe University of Applied Sciences, Germany
May '22 - Aug '22
Research funded by the DAAD WISE Scholarship [Code][Website]
Designed an end to end pipeline for multi-view stereo dense 3D reconstruction from
a handheld stereo-camera(Intel RealSense) that outputs stable dense pointclouds. Integration of classical visual SLAM algorithms with U-Net adapted deep learning architectures for dense depth prediction.
Researcher | APPCAIR & Intel Labs
May '22 - Aug '22
Demonstrated use of ensemble learning for the task of activity recognition/video classifi-
cation on the Something-Something-v2 dataset. Weak learners were typically used
for feature extraction. Explored various methods for combination of features, ultimately
used for downstream classification
Machine Learning Research Intern | CEERI Pilani
May '21 - Sept '21
Worked under under the supervision of Prof. Sandeep Joshi
and Prof. Madan Lakshmanan.
on classification of a person as fatigued or non-fatigued based on PPG signals of a human subject.
This template is a modification to Jon Barron's website. Last Updated: January 2025.