Master of Science (MS)・
Computer Engineering 2017 - 2020
Activities and societies
Elected Student Parliamentarian. Responsibility includes:
- student semester tickets
- electing and controlling the ASTa
- deciding the budgeting of the student bodies
Description
Major focus: Intelligent systems and Data Science
Embedded Machine Learning: 1-year long Project Group.
3 Research / Survey papers in Machine Learning area.
Projects:
1. Master Thesis - Deep learning for anomaly detection in autonomous systems
Realized state-of-the-art deep learning models to carry out anomaly detection on time-series data from the autonomous race car. Perform hyperparameter optimization, neural architecture search, unsupervised feature selection and create interactive visualisations
2. Predictive Maintenance in Autonomous Systems
Incorporation of ML algorithms for anomaly detection in an Industry 4.0 project - GMMs, Clustering, Decision Trees, ARIMA, etc
3. Embedded Machine Learning
Predictive maintenance for real-life industrial use-cases. Compressed the models and deployed it on Xilinx PYNQ board.
4. Sparsity in Neural Networks
Research on sparsity algorithms to prune and shrink large neural network models and run them on embedded platforms.
5. Face Recognition using Convolutional Neural Networks (CNNs)
Constructed and evaluated the state of the art models for face detection and recognition - LeNet, AlexNet, VGG, GoogLeNet, ResNet.
6. Chatbot using Natural Language Toolkit (NLTK)
Fabricated a naive chatbot trained on the Wikipedia corpus to respond to the user queries.
7. Face Generator using Generative Adversarial Networks (GANs)
Deep Convolutional GAN to generate realistic looking faces. Trained on the celebrities face dataset.
8. Neural Style Transfer
Compose images in the style of the artistic image using deep learning model.
9. Detection of Malicious URLs using Autoencoders
Realized a URL classification model design using Deep Autoencoder and filter suspicious URLs.
10. Credit Card Fraud Detection and Prevention using Deep Learning (DL)
Deep autoencoder trained on the Kaggle Finance dataset for credit card transactions use-case.
11. Sentiment Analysis of Trip Advisor Reviews: Natural Language Processing (NLP)
Used Bag of words model to analyze the sentiments in the feedback provided by the customers.
12. Facial Emotion Recognition: Inception Network and other popular CNNs
Face detection followed by emotion recognition using CNNs on the bounded face detected using OpenCV.
13. Object Recognition on Android: Tensorflow Lite Models
The model classifies the objects shown on the android device’s rear camera in real-time.
14. Simulating Self Driving Car: Udacity Nanodegree
Engaged with computer vision and deep learning to automotive problems, like detecting lane lines, predicting steering angles, etc.
15. Verilog HDL Compiler: Computer Science Project
Devised a compiler to achieve the lexical, syntactical, and semantic analysis of a Verilog HDL code.