Darren Ramsook
Hi! My name is Darren Ramsook & I’m a PhD Candidate at Trinity College Dublin (under the Sigmedia Group), working on perceptually optimized video restoration (a mixture of video processing and deep learning). I have a with a keen interest in image/video signal processing. This interest in signal processing is mainly due to the arts having a special place in my heart & also because I’m an avid musician. I have my MSc. Data Science, BSc. Electrical & Computer Engineering from the University of the West Indies. I combine my love for signal processing, software engineering and data science to help systems better analyze the world around us.
Skills
-
Python |
C++ |
MATLAB |
Tensorflow |
Keras |
scikit-learn |
pandas |
bokeh |
dash/plotly |
SQL/NoSQL |
flask |
- - Image/Video and Audio Signal Analysis
- - Machine Learning and A.I. based predictions
- - Deep Neural Networks
- - Data Visualization and Web Development
- - Profieicency in many Python Packages – numpy, scipy, flask, matplotlib, Tensorflow, etc.
- - Strong Analytical/Mathematical Skills in Algebra, Statistics and Probability
Experience
PhD. Student
- Research into perceptually optimized video restoration for video signals.
- Used deep convolutional neural networks to create a differentiable estimate of a complex non-diffrentiable perceptual visual quality metric (VMAF) for use in optimization scenarios.
- Created spatio-temporal denoiser networks that are trained using the differentiable estimate of VMAF in the spatial (image) and temporal domain (video).
Intern (Research/Engineer)
- Utilized Machine Learning and Deep Learning to optimize various imaging processes.
Adjunct Research Scientist
- Research into abstractive Textual Summarization for Instant Messages on devices with limited screen space (eg. smartwatches).
- Created both a Transformer Neural Network and a Greedy N-gram token based optimization approach to solve this problem.
- Performed qualitative tests on the improvement in both the reading and understanding time of the summarized versions of the problem.
Data Scientist Intern
- Formulation of a single figure of metric for evaluating real time network performance and customer experience for 2G (GSM), 3G (UMTS) and 4G (LTE) Mobile Networks. This required indepth end-to-end data science skills such as - domain upskill, data processing and cleaning, mathematical formulation and end-to-end deployment.
- Evaluating Customer Feedback over Social Media in clustered geographic areas autonomously using sentiment analysis.
- Weekly workshops on Programming, Mathematics, and the entire Data Science Project pipeline (ETL, Model Creation & Testing, Deployment)
Associate Professional - Research
- Research and design/development of classification and regression trees (CART) Hybrid Intrusion Detection System in Python that performs classification based on user activity and geolocation metadata with an accuracy of 94.5%(Jaccard Similarlity).
- Conducted Network Threat Analysis for educational domains utilizing by mimicking conventional web access points (HTTP, SSH, FTP) and an Industrial Control System(ICS).
- Research and development in the creation of a real time audio equalization system through the use of image and acoustic feedback of a given environment in collaboration with Indigisounds Inc.
- Assisted with the deliverance of main project for the Introduction To Programming Course (ECNG1009).
Associate Professional - Software
- Worked within the Data Analytics team in the Trinidad Microsoft Innovation Center.
- Created an automated print media system that replaced internal paper-based process for approval of company graphics.
- Installed and integrated data acquisition sensors into an existing solution for refrigerant systems in a major retail company.
Systems Engineer Intern
- Worked with Government level clients to design an automated audio/visual system for use in Judicial & Banking Environments.
- Oversaw & completed system implementation of two designed systems in an efficient and on-time manner.
- Designed,Programmed and Deployed, in Python, an Interactive Historical Cash Register System for use in the Central Bank of Trinidad & Tobago.
Education
Trinity College Dublin
University of the West Indies
University of the West Indies
Academic Papers
Perceptually motivated deep neural network for video compression artifact removal
Figure of Merit for a Multi-Generation Network
A differentiable VMAF proxy as a loss function for video noise reduction
A differentiable estimator of VMAF for video
Cybersecurity Threat Analysis for an Energy Rich, Small Island Developing State
Intrusion Detection and Avoidance for a Heterogeneous Cluster of Web Sites
Projects
Projects that I've worked on:
A differentiable VMAF proxy as a loss function for video noise reduction
Perceptually optimized video restoration.
Traditional metrics for evaluating video quality do not completely capture the nuances of the Human Visual System (HVS), however they are simple to use for quantitatively optimizing parameters in enhancement or restoration. Modern Full-Reference Perceptual Visual Quality Metrics (PVQMs) such as the video multi-method assessment fusion (VMAF) function are more robust than traditional metrics in terms of the HVS, but they are generally complex and non-differentiable. This lack of differentiability means that they cannot be readily used in optimization scenarios for enhancement or restoration. In this paper we look at the formulation of a perceptually motivated restoration framework for video. We deploy this process in the context of denoising by training a spatio-temporal denoiser deep convultional neural network (DCNN).
Video Restoration Human Visual System Perceptual Optimization Deep Convolutional Neural Networks Tensorflow/Keras
A differentiable estimator of VMAF for Video
Making subjective PVQMs differentaible.
There has been limited use of perceptual visual quality metrics for automated image quality optimisation in applications like video enhancement. This is because many useful visual quality metrics for video e.g. VMAF are complex and non-differentiable. This work proposes the use of a CNN architecture to approximate the temporal behaviour of VMAF. Employing degradation generated with H.265 compression, our model achieves a 4.41% RMSE in predicting VMAF. This can now be deployed as a video based loss function in video enhancement and compression tasks.
Perceptual Optimization Video Processing Deep Convolutional Neural Networks Tensorflow/Keras
Instant Message Summarization with Emoji Unicode Characterset Support
Abstractive Susmmarization of Instant Messages
Research into two models for summarizing an instant message, through the use of text and the Emoji Unicode Characterset, for applications where screen space is limited (eg smartwatches). The first proposed model utilizes a Greedy N-gram token replacement method, while the second proposed model utilizes a Transformer network. The Transformer network was trained on a set of 532 manually labelled instant messages and 92554 automatically labelled dialogue sentences. Initial results showed the Greedy N-gram token replacement method produced higher quality results and this method was evluated using human participants. It was found that there is an increase in average time taken to read and interpret the summarized messages when compared to the unsummarized messages.
Summarization Abstractive Summarization Extractive Summarization Transformer
Realtime Network Performance and Customer Experience
Mathematical Framework done for TSTT
Developed and implemented a framework for determination of a single Figure of Merit (FoM) that can be used for high level monitoring while at the same time providing sufficiently valuable low level indicators to assist with the isolation and detection of problems. This was illustrated using data from a real cellular network.
Figure of Merit Performance Monitor Data Analytics Key Performance Indicator Telecommunications Infrastructure
Web Server Instrusion Detection and Avoidance
Trained A.I. Based System for Web Server Instrusion Detection and Avoidance
This program utilizes the combined access log format of apache web servers to extract certain key features of actions and classifies users in one of three possible groups (Malicious, Unknown & Safe). This system allows for a headless operation once deployed. The system automatically performs pre-processing, classification and populates a list of Internet Protocol addresses whose access to the web server is prohibited. This system was made open source for both public use and development. This trained model resulted in an accuracy score of 94.5% on the test set.
Machine Learning Classification & Regression Trees (CART) Big Data SKLearn Python Software Development
ID3 Split Calculator
Decision Tree Split Calculator
A decision tree learning calculator for the Iterative Dichotomiser 3 (ID3) algorithm. By utilizing the ID3 Algorithm, the best feature to split on is decided. This program requires to additional libraries outside of the default libraries included with Python (math, csv). Therefore this needs to extra set-up configuration. Tested and working on Python 3.7.
Python Optimization Decision Tree ID3
Social Media Impression Calculator
Optimal Impression Calculator for Multi-Stage Advertisement Campaigns on Social Media
This calculator was built based on the assumption that within a given Social Network, friends of a given person has an effect on the probability of that person clicking an advertisement. This calculator evaluates a scenario where advertisements are split up between two stages, with the second stage being affected by what has occured in the first stage. For more information about this check out this paper - On the Problem of Multi-Staged Impression Allocation in Online Social Networks, done by Inzamam Rahaman and Patrick Hosein.
Optimization Operations Research Big Data Algorithms Probablistic Modelling Python Tkinter
Cash Register System
Interactive Historic Cash Register System for the Central Bank Mueseum of Trinidad & Tobago
This cash register was designed, created and deployed for use in the Central Bank Mueseum of Trinidad & Tobago. It allows the user to scan items located in the mueseum. The user would then be presented with information about the scan as well as the prices across various time periods. It was created with the main user base being younger children.
Software Engineering Python UI/UX Design Interactive Electrical System Raspberry Pi
MicroController Based Pulse & Glucose Monitor
Medical Device
The created system was able to detect the heart-rate & glucose level of an individual, store the data, and give out an alarm if the heart rate drops below a certain threshold. A keypad & LCD display was used as means of system interaction.
Microcontroller PIC C Sensor Design Circuit Design Testing/Analysis
Certifications
-
Research Integrity - Engineering & Technology
EpigeumFebruary 2021 -
Fundamentals of Accelerated Computing with (CUDA C/C++, Python)
Nvidia Deep Learning InstituteAugust 2019 -
Deep Learning Specialization
CourseraJuly 2019 -
Machine Learning
CourseraOctober 2018 -
Cybersecurity Fundamentals Certificate (CSX)
ISACAAugust 2017 - Awarded a full Scholarship through the D-Real program to pursue PhD Electronic & Electrical Engineering at Trinity College Dublin.
- Created Notebooks for the Nvidia backed RAPIDS Open Source Library.
- Scholarship in the field of Mathematics due to academic performance in the Caribbean Advanced Proficiency Examination (Diploma) in the subjects Pure Mathematics, Applied Mathematics, Physics & I.T.; granted by the Government of Trinidad & Tobago.
- Placed 8th in the Caribbean Region for the I.T. Advanced Level (CAPE) Examinations (2015).
- Trinidad & Tobago National Music Festival 1st place two-time winner.
- Fully funded scholarship for obtaining the Cybersecurity Fundamentals (CSX) Certificate from the Government of Trinidad & Tobago.
Achievements
❤️ Vali