This internship was about implementation of the Digital Audio Processing Dialect (DAP) using MLIR to improvise the performance of the audio processing filters such as Biquad and IIR filters.

The objective of this project was to program an FPGA based robot using Haskell.

My project was Ball levitation using computer vision on FPGA.

My project was about replicating Google’s implementation of the Paper titled “ Accelerating eye movement research via accurate and affordable smartphone eye tracking”.

My internship was about making a tool in python to capture IPv6 packets as a system service and develop a debian package for the same.

The work is not complete yet. As of now, I am spending time to become more familiar with their custom infrastructure for domain-extensible compilers.

Worked on GPGPU co-scheduling algorithm.

The goal of this project was to develop a test platform for RISC-V processors on FPGAs dedicated to artificial intelligence at the edge.

My project is about the Senders and Receivers Proposal P2300 of the ISO C++ Community which is about to be accepted into C++26. It is a generalized programming model for synchronous and asynchronous programming.

Worked on developing firmware for and synchronize real-time multi sensor stream data.

I am yet to complete it. But I plan to focus on getting papers out of it.

The goal of my project, "SBR: Self balancing bot," was to build an effective and modular CAD design of a self balancing bot that would be used as an educational product for various personas.

Adding features and more compatibility for simpPRU was the main aim of the project.

My project was about creating an algorithm to perform convolutions on fully memory-distributed data.

The project I worked on was Improving the gstreamer element to add support for properties.

My project was about improving and tweaking the performance of the RADIS codebase.

I had to develop an autonomous nano drone stack, and perform indoor localisation on it.

Create prototype for Contact Sensing Foot, improved design of original Foot, designed heat sink, solving overheating problems.

Creating multiple prototypes of a responsive foot, current quadruped at the lab(Stoch-3) had some structural failures and parts used to bend after operation. So my task was to create a closed structure for abduction (abduction coupling).

Worked on link prediction using Graph Neural Networks.

The work involved creation of a novel algorithm with custom boundary extrapolation, specifically suited for wider vector architectures such as RISC-V.

We are developing a performance oriented C++ library for Image processing using MLIR as its backend.

It was about developing a Python API in C++. Later the major focus has been their installer script.

Worked on simulating embedded systems before deploying them using DEVS formalism.

Worked on adding multi-stream support to the gstreamer element.

Designed and Developed gateware for BeagleWire FPGA Cape.

Designed and developed a FPGA framework to verify the SHAKTI processors on the FPGAs

Implemented Bela on BeagleBoard and Developed a standardized setup that allows an easy jump across all TI chips.

Worked on the Software Aspects of the 5 Dof Robotic Arm and also created the online interfaces to teleoperate the Arm from any city in the world

Developed a Deep Learning System that can detect and count the number of strawberries (ripe, unripe) in the strawberry field and then predict the yield accordingly

Added support for spatial datasets to the retrieverdash project

Redesigned and Optimised GIL & 2D Convolution/Correlation APIs

Developed and deployed a Block-Based TL-Verilog solution for developers and new entrants to the sphere

Implemented Beagle-Config on BeagleBone Boards

Implemented BMQ(Bit Map Queue) and PDS-mq(Priority and Deadline based Skiplist multiple queue) cpu scheduler support for Linux Kernel 5.10

Worked on creating a transpiled language called simpPRU

Developed an application for vehicle counting, detecting violations like wrong side, red light, speed, helmet and triple seat.

Developed an application for vehicle counting, detecting violations like wrong side, red light, speed, helmet and triple seat.

Trained a model which can detect, classify and localise Indian vehicles.

Collected and Annotated large datasets for custom object detection such as helmets, guns.

Automated the process of calibration to reduce time and errors

Tried to implement ‘Time Crystals’ as a scalar field for inflationary cosmology