Hi,
I am Suryakiran
Sureshkumar

Download Resume

About

I am Suryakiran Sureshkumar

I'm currently pursuing Master of Science in Computer Science at NYU's Courant Institute of Mathematical Sciences, specializing in advanced topics like machine learning, data science, and software development. Prior to this, I served as a Machine Learning Engineer at Tiger Analytics, leading projects in computer vision and machine learning for global clients and overseeing a team managing a no-code data science platform. I've also honed skills in cloud computing, natural language processing, and data analytics through hackathons and projects. My passion lies in tackling complex challenges with data and technology, always eager to acquire new skills and tools. I prioritize collaboration, innovation, and quality in my work, aiming to deliver solutions that meet client and stakeholder expectations. I'm open to new opportunities and connections in the realms of machine learning, data science, and software engineering.

Skills

Technical Skills

Programming Languages: Python, C++, C, Java, Go

Web Technologies: HTML, CSS, JavaScript, PHP, React.js, Node.js, Express.js, jQuery

Database: MySQL, Oracle, MongoDB, PostgreSQL

Cloud Technologies: Azure, Oracle

Tools & Platforms: Postman, Power Apps, Tableau, DataBricks, Docker, GitHub, GitLab, Azure DevOps

Frameworks: Flask, Django, FastAPI, Flutter, Spark, PyTorch, Tensorflow

Libraries: Pandas, NumPy, Scikit-Learn, OpenCV, PySpark, Matplotlib

Projects

Project Image

DriveVLM

Developed a novel method to tackle supervised data scarcity in autonomous driving by fine-tuning the QwenVL Vision-Language Model. Tested on CARLA driving routes, our method processed visual inputs to generate precise vehicle control commands. With just 5% of the data, we achieved 40% of the traditional models' performance, closely matching them at 20%, and surpassing at 40%, showcasing the potential of Vision-Language Models in enhancing data efficiency.

GitHub
Project Image

ChatLoom

It is a specialized chatbot leveraging OpenAI's advanced language models, ChatLoom is tailored to provide detailed answers on cosmology and astrophysics topics. With a focus on replicating the expertise found in the cosmos field, this chatbot offers a seamless local setup, making cosmic knowledge accessible right from your desktop.

GitHub
Project Image

SegMentor

It is a novel future frame segmentation approach using the SimVP model and MogaNet. It predicts the 22nd segmented frame using DeepLabV3+ network from 11 RGB frames, achieving a Jaccard Index score of 0.2952 on validation and 0.2852 on the hidden test set, demonstrating its potential for applications in computer vision, including Autonomous Driving.

GitHub
Project Image

Pokemon GAN

Created a DC-GAN (Deep Convolutional Generative Adversarial Network) using PyTorch programming language and trained it on a dataset of Pokemon images. This allowed me to generate brand new Pokemon images that had not been seen before.

GitHub
Project Image

Using Video Summarization For Effective Search

This system improves the searchability and indexing of videos in a database using techniques such as video summarization, keyframe extraction, image captioning, and removing common words (stop words). It is implemented using Pytorch and extracts relevant keywords for input videos to improve search results.

GitHub
Project Image

Deep Learning Framework for Component Identification

It is a setup to monitor the assembly line of a manufacturing unit and use object detection algorithm to detect, classify and count the various components in the monitored image. It also extracts the coordinates of various components in the assembly line and provide it to an automated system.

Paper
Project Image

Covid Fight

This is a mobile application that helps users during the coronavirus pandemic. The features of the app include delivering a live count of cases, a crisis communication chatbot that helps users resolve their doubts regarding the pandemic, a self-diagnosis feature, notifying the users to wash their hands regularly, and giving them immediate access to the distress number.

GitHub
Project Image

Phishing Site Detection

It is a web application with a chrome extension that detects whether a particular website is phishing or not, using a neural network, by extracting 14 features from the particular site. A comprehensive analysis of the attack will be periodically sent to the cyber department.

GitHub
Project Image

Night-time Object Detection using Machine Learning

It is a setup which uses CCTV camera and detects required objects in night. We used a modified YOLO algorithm exclusively for object detection during night time .We presented paper on this project in the Seventh National Conference on Computer Vision, Pattern Recognition, Image Processing and Graphics.

Paper

Contact

Hello! Ask anything about Surya.
Chatbot Icon
×