Svs Dhanush

Who Am I?

A software engineer @ SONY India Software Centre (SISC), Bengaluru. My curiosity thrives on how technology molds our world, and eager to be a part of it. Love problem solving and turning ideas into reality. An amateur photographer who see things from different perspective that others might overlook.

2018 - 2020

Sri Chaitanya Junior College

Board of Intermediate Education A.P (BIE AP)

2020 - 2024

Amrita Vishwa Vidyapeetham

Computer Science - AI

2024 - present

SONY India Software Centre

Software Engineer

C++
Python
MATLAB
SQL

PyTorch
TensorFlow

Flutter
Firebase
Android Studio

MySQL
MangoDB

Research

An Empirical Comparative Study of Machine Learning Algorithms for Telugu News Classification. (Accepted)

Leveraging NLP techniques, this study employs classification techniques by developing different ML and DL models, utilizing various word embeddings, aiming to enhance efficiency in categorizing newspaper articles.

A Comparative Study of Text-To-Speech (TTS) Models and Vocoder Combinations for High-Quality Synthesized Speech. (Accepted)

The paper presents a comprehensive comparison of multiple Text-to-Speech (TTS) models and vocoders. The primary objective is to identify the most effective TTS model-vocoder combinations.

Detection of Mental illness from social media Text (Accepted)

Explore various methods to enhance the accuracy of detecting different mental illnesses from social media text (such as Reddit). The primary method used involves the ensemble embedding of two different transformer models.

HousieTambola Numbers Generate

The HousieTambola Numbers Generator is an all-in-one app for enhancing your Tambola/Housie game nights. It offers features like random number generation, multilingual voice calls, and a detailed notes section etc.

Available now on the Play Store.

Play Store >

Projects

Travelers

A mobile application enabling users to connect with other travelers who are traveling to the same destination. Build using Flutter and Firebase.

Visit Travelers App >

Deep Fake Video Detection

Developed a Deep Learning based model to classify whether a video is real or deepfake using ResNeXT50(CNN) and LSTM.

GitHub >

Image Editing App

A Flutter based mobile application to perform image enhancement and transformation tasks. Supports both Android and iOS.

GitHub >

Expense Tracker

An expense tracker to track expenses from the invoice image using AWS. Uses Amazon Textract Service to scan and extract text from the invoice image and S3 Bucket for data storage.

GitHub >

Smart Home Automation

Developed a interactive python program that seamlessly executes tasks based on both voice and message commands using Raspberry Pi. Enables message based communication via Telegram app.

GitHub >

Amrita Navigator

A user-friendly map interface application for our college that enables users to find the shortest path between a given source and destination uses.

GitHub >

Ball Chaser using ROS

Using the Robotics Operating system (ROS) capability and gazebo platform to construct a basic ball tracking vehicle. Our goal was to create a rudimentary prototype for a bot that can detect and follow color.

GitHub >

Pulmonary nodule classification

Classify the pulmonary nodules using SVM and KNN algorithms trained on LIDC-IDRI dataset and its subset LUNA16.

GitHub >

Coal Mine Surveillance Bot using ThinkerCAD

A autonomous bot, to detect the temperature level, fires to prevent major accidents and can detect any human presence at the restricted, accident areas.

GitHub >