Jan 2020 – Feb 2020 · 1 mo · Singapore

• Devising a truly subquadratic algorithm for clustering reads in DNA storage, known to have many small clusters in edit distance space.

Dec 2019 – Jan 2020 · 1 mo · Jeddah Governorate, Saudi Arabia

• Worked towards a profiling technique for deep neural network Optimizer
• Implemented Q-Learning for optimization of hyperparameters built a modular code which can be used to install the profiling technique in any given optimizer.
• Employed and experimented for multiple DNNs for multiple datasets on the HPC.

Aug 2019 – Nov 2019 · 3 mos · New Delhi, Delhi, India

• Worked as a teaching assistant for the course Introduction to Parallel and Distributed Programming(CSL380), IIT Jammu

May 2019 – Jul 2019 · 2 mos · Bangalore, India

• Worked in Java for deployment of Android Application for the Samsung One, IOT Division.
• Trained ML Model(transfer learning in fasttext), for text classification and device prediction.
• Employed heuristics and image processing to remove Ad blocks from mobile screenshots.
• Worked with Tesseract and Firebase OCRs to extract text data from processed screenshots.
• Explored the SGNNs, to build a computationally light ML model for classification in mobiles.

Jul 2018 – Aug 2018 · 1 mo · New Delhi Area, India

• Devised an identity framework, presented white paper on Self Sovereign Identity framework of Appspicket startup.
• Involves the use of Cryptographic constructs :- Zero Knowledge Proofs, Anonymous Credentials and Blind signatures.
• Incorporating a factor of Social trust into the framework to solve the problems the Adhaar framework is facing in India.

May 2018 – Jul 2018 · 2 mos · New Delhi Area, India

• Worked to predict when a human is going to speak using muted video(regular camera) inputs of chest movement(manually collected).
• Extracted respiration data from videos using Optical Flow Vector computation in a region with pixels moving in sync.
• Trained LSTM Deep Learning Model to accurately detect the presence of speech transition in a respiration window.
• Computationally efficient method is employed for detection of contact free respiration patterns, highly suitable for mobile.

Dec 2017 – Jan 2018 · 1 mo · New Delhi Area, India

• Encoded VPPM and VOOK modulation schemes using the concept of Dimming, in VHDL, implemented on FPGA.
• Implemented using Differential Modulation of the signal to properly synchronise the receiver and align its local clock.
• Developed a model which provides multiple modulation schemes and data transfer rates to select from to transmit data.