Portfolio Category

Communications

Showing 5 Result(s)

Speech Denoising with Deep Nets

on

In this project I trained neural networks using MATLAB to remove the noise from speech signals with background noise to enhance the quality and the intelligibility of the speech.

For this, I created two pipelines. The first pipeline uses the Discrete Cosine Transform (DCT) and the second pipeline uses the Short-Time Fourier Transform (STFT). Their diagrams can be found below.

I ran multiple tests for both pipelines using different parameters such as the window length, the window type, the overlap, the neural network type, and the number of network layers.

To measure the performance and identify the best network, I calculated the average MSE between the original and the denoised signals across all test files. The table below summarizes the performance of the system for each test. The best results gave an MSE of 0.0029!

Listen to the original (top), noisy (middle), and denoised (bottom) signals below, and check out their spectrograms!




Check out the slide deck and send me a message to learn more!

Mediacasting Analysis

on

This project consists of performing an in-depth analysis of the main technologies which are currently used by companies to provide content delivery to their customers, as well as to explore the Weighted Slope One method for predicting the ratings that customers are likely to give to a range of different movies. The first part of the analysis includes Content Distribution Networks (CDNs), Peer to Peer (P2P) Networks and Hybrid CDN-P2P networks used for content delivery. Additionally, the second part of the analysis focuses on the Weighted Slope One method as a Collaborative Filtering approach to predict user ratings for media contents.

I completed this project at the University of Surrey, UK. For more information on the project click on the link to the repository.

Project Repository

4-PAM Digital Communication System

on

This project consists of a simulation of a Digital Communication Systems (DCS) vector model which implements the 4-PAM (Pulse Amplitude Modulation) technique using MATLAB. In it, I compared the probability of bit error as a function of the signal-to-noise ratio per bit of the simulated system to the theoretical values. The detector is assumed to be a single correlation-type demodulator.

I completed this project at the University of Surrey, UK. For more information on the project click on the link to the repository.

Project Repository

FM Receiver

on

The objective of this project was to build a prototype of an FM radio receiver using an FM chip (AR1010) and an audio amplifier (TDA2822). The FM receiver is controlled using a PIC. The receiver is to be tuneable approximately over UK broadcast band (87.5 – 108 MHz) and produce a stereo output. It should also be able to drive a set of 35 ohm headphones and be powered by 3AA batteries (4.5V).

I worked in a team to complete this project at the University of Surrey, UK.

For more information on the hardware and firmware specifics of the project email me to get access to the repository!

Simple Chatbot

on

This weekend-long lab project consists of a simple bilingual chatbot that allows people to take a survey for a sightseeing tour. It uses a Raspberry Pi. I completed this project at Cornell Tech.

Check out the videos below! For more information click on the link to the project repository.

Project Repository