Aug 2021 – Feb 2026 · 4 yrs 6 mos

• Projects:
• 1) Lidar Camera Calibration
• Contributed to the development of the algorithm to perform target based calibration between the lidar and camera sensor.
• Closely worked with UX Researcher and Designer to resolve the various usability issues and reduce the bug count by 50% and with the Lidar Camera Calibrator App.
• Skills: Computer Vision · Architecture Patterns · MATLAB
• 2) Determine Safe Landing Zones for Aerial Vehicles
• Designed and implemented a statistical feature-based segmentation algorithm, to segment the point cloud into four classes: Dangerous, Unsuitable, Risky, and Suitable landing zones.
• Optimized point cloud processing through custom overlapping blocks, resulting in a remarkable 3x runtime improvement compared to standard processing methods while mapping unknown environments using Lidar-SLAM on UAV/Drones, simulated in Unreal Engine.
• Skills: Computer Vision · SLAM · Machine Learning · MATLAB · Python · Research
• 3) Curb Detection and Tracking in 3-D Lidar Point Cloud
• Developed a feature-based curb segmentation algorithm employing Kalman Filters to track curb points across multiple frames which enables path planning and safe navigation in autonomous driving scenarios where GPS signals are unavailable.
• Skills: Kalman Filters · Computer Vision · MATLAB · C++ · ROS · Research · API Development
• Leadership Skills:
• Experienced in skilfully navigating negotiations across diverse stakeholders, multi-functional teams, and international contexts.
• Demonstrated leadership in customer engagement by proactively collaborating with prominent organisations, effectively identifying and resolving challenges while ensuring the smooth implementation of workflow solutions.
• Prepared and delivered concise presentations summarising conference papers from events such as CVPR.

Jul 2021 – Jul 2021 · 0 mo

• Point Cloud Processing
• Developed a custom algorithm to convert 3-D point clouds to 2-D lidar scans.
• Designed and developed an algorithm to convert unorganized point clouds to organized point clouds. This functionality is used extensively across various segmentation workflows and in undistorting ego motion.
• Skills: SOLID principles · API Development · MATLAB · C++
• Other Deliverables:
• Optimizing the performance of multiple APIs, resulting in a remarkable up-to 11 times faster runtime.
• Crafted unit tests that achieved a code coverage exceeding 90%.
• Updated 60% of the existing codebase to adhere to internal coding guidelines.

May 2020 – May 2021 · 1 yr · Hyderabad, Telangana, India

• 1) Terrain Classification On Aerial Lidar Data
• Developed and deployed feature based algorithm to classify the input point clouds into: Ground, Buildings and Vegetation, using Region Growing based Segmentation algorithm utilising normal and curvature values.
• Conducted an in-depth literature review and prototyped ground segmentation functions.
• Skills: Machine Learning · Graph algorithms · Research · MATLAB
• 2) Aerial Lidar Semantic Segmentation
• Leveraged the PointNet++ Deep Learning network to segment the input point cloud into 8 classes, namely, Ground, Building, Vegetation, Powerlines, Poles, Fences, Cars, and Trucks.
• Pioneered an efficient training strategy, reducing training time upto 50% on the same GPU hardware.
• Achieved overall accuracy and IoU exceeding 80% for critical classes like Ground, Vegetation, Building, and Powerlines.
• Skills: Deep Learning · MATLAB · Python

Jul 2019 – Jul 2019 · 0 mo · Hyderabad Area, India

• Among the top 50 to be shortlisted for Summer School on Computer Vision at IIIT Hyderabad.
• Learned about various projects using CNN, Semantic Segmentation, GAN
• Worked on Sequence generator using RNN, CT scan Medical Images

Dec 2018 – Jan 2020 · 1 yr 1 mo · Chennai Area, India

• AgriMax vision is to drive hyper-local urban food production. We make farming so much simple and easy that now anybody can grow their own food.
• My Roles:
• Pre Incubated 4Climate Technologies to venture into Hydroponic Food Computer space.
• As the Tech-lead, researched and developed CNN models for disease classification in plants due to nutrient deficiency based on spots, leaf color and pigmentation.
• Developed an expert system model that can predict best crops to be grown.
• Redesigned the system by integrating IoT and Android app to monitor plant growth parameters.
• Deployed the app successfully on PlayStore
• Developed Product Strategy & Analyzed market to identify follow-on markets.

May 2018 – Jul 2018 · 2 mos · Chennai Area, India

• R.I.S.E. Group is a research group in the CSE Dept. of IIT Madras working in the areas of Computer Architecture, Security, Machine Learning and VLSI Design. Shakti microprocessor which is based on RISCV architecture. Poky is an kernel based on Open Embedded Yocto project. Caffe is a Deep Learning Framework, developed by UC Berkley, used mainly for image classification.
• My Roles:
• Researched on Computer Architecture, Security, Machine Learning and VLSI Design at RISE Labs
• Responsible for Porting Caffe (Deep Learning Framework) to Shakti Processor (RISCV based) using Poky Kernel.

Jun 2017 – Mar 2018 · 9 mos · Chennai Area, India

• A funded project at Design Innovation Center (by MHRD) to develop a working prototype of KIES, design of Smart Lock as part of Home Automation and Security.
• My Roles:-
• Developed a Smart Door Lock prototype with Face Recognition based Authentication using SVM.
• Developed encrypted password sharing authentication system which has 2 layers of security.
• Enabled encrypted password sharing, so that others can access your door.
• Drive Link : https://drive.google.com/drive/folders/1oT_LhdUdKi_FQu2Bdl8enqZQI5lb2yCj?usp=sharing