About Me

I am Sargam Goyal, a pre-final year undergraduate student at IIT Roorkee pursuing B.Tech in Data Science and Artificial Intelligence. I am an active member of the Data Science Group at IIT Roorkee and have served as an Undergraduate Teaching Assistant for courses like C++ and Digital Electronics. Additionally, I contribute to the student technical magazine, Geek Gazette, as an editorial member. I had the opportunity to be a research intern at the CrIS Lab, IISc Bengaluru during summer 2025.I have academic interests in AI security and computer vision applications, particularly autonomous systems. Currently, I am finding agentic AI interesting which I am exploring in my free time.


Projects

Project 1 Screenshot

EfficientNet-Based LULC Classification using PlanetScope and Sentinel-1 Data

A comparative analysis of supervised LULC classifier using PlanetScope and Sentinel-1 satellite imagery

Tech: PyTorch, Docker, SNAP, QGIS, EfficientNet

Result: Concluded that PlanetScope data performed better despite lesser bands

GitHub Repo

Project 2 Screenshot

Design and Optimization of CNN-Based Facial Emotion Recognition System

A facial expression recognition models using the FER2013 dataset

Tech: PyTorch, CNN, SVM, Logistic Regression, Random Forest

Result: Achieved a test accuracy of 53%, surpassing 81.54% of benchmark human-level performance on the dataset.

GitHub Repo

Project 3 Screenshot

Implementation of YOLOv9 on Self Driving Car Dataset

A YOLOv9 object detector to enhance self-driving car perception in bad weather, targeting improved safety in adverse conditions.

Tech: Ultralytics, labelImg

Result:Achieved 0.595 mAP@0.5, competitive with leading benchmarks (0.50-0.95 mAP@0.5).

GitHub Repo