Machine Learning Research
Continuous Speech Recognition Sr.
Full-Stack Web Developer Jr.
ABOUT ME
Hello there! I'm Christian a machine learning Scientist with background in continuous speech recognition and images processing. I was fortunate to be supported by a
PUC-Rio master and doctoral Fellowship, which gave me the opportunity to trek around the world showing my principal work of enhancement and speech recognition.
In my 7 years as a research developed strong communication skills from extensive talks.
I have experience and ability to work independently or as part of a team particulary in the following areas: Machine/Deep Learning, DSP technology applied to Speech Recognition,
Audio Processing, Noise Reduction, Noise Suppression, Multispectral and Hyperspectral Image Reconstruction, and also I'm a Principal Reviewer of Journal of Electrical and Electronic Engineering.
I have published over 8 articles an gave over 5 public talks and lectures and I even achieved great recognition which I feel proud because I got a mention as the best conference paper in the Continuous Voice Enhancement and Processing
FORMAL EDUCATION / DEGREE
Pontifical Catholic University of Rio de Janeiro. Brazil
Degree: Ph.D. in Electronic Engineering
Thesis: Enhancement and continuous speech recognition in adverse environments
February 2014 – April 2018
Pontifical Catholic University of Rio de Janeiro. Brazil
Degree: M.Sc in Electronic Engineering
Dissertation: Continuous Speech Recognition with MFCC and PNCC features and robustness techniques Wavelet Denoising, Histogram Mapping, Spectral Subtraction and Filter with Neural Networks
February 2011 – November 2013
Francisco de Paula Santander University. Colombia
Degree: B.S. Electronic Engineer
Dissertation: Supervise and support in mechatronic automation projects of the International Clean Production Center Lope from SENA
January 2005 – February 2011
EXPERIENCE
MACHINE LEARNING ENGINEER LEAD
VOZY
Sep 2021 - present
Leadership, scientific research and software development
Lead the team, research and development of new machine learning techniques so that conversational agents
can generate a better conversational flow, improving the acoustic and language models used for NLP, STT and TTS architectures.
Technologies used: Python, TensorFlow, Keras, Pytorch.
MACHINE LEARNING ENGINEER
VOZY
Sep 2020 - Sep 2021
Scientific research and software development
Development and research in the area of speech recognition and audio processing for new and challenging
environments using advanced machine learning techniques.
Technologies used: Python, TensorFlow, Keras, Pytorch.
POSTDOCTORAL RESEARCH FELLOW
High Dimensional Signal Processing Reseach Group (HDSP)
May 2020 - Present
Research Scientist
Development of an optical-computational system for the fusion of depth and hyperspectral images using deep learning techniques
and their application in the classification of citrus by its level of maturity.
Technologies used: Python, TensorFlow, Keras.
INDEPENDENT EXPERIENCE
Sole Researcher
Apr 2019 - Apr 2020
Research Scientist
Investigate and analyze the problem presented by people with hearing disabilities and pathogenic diseases such as ALS, in order to develop a continuous voice recognition system
that can mitigate your communication problem in real-time using artificial intelligence techniques. The system is being created for Latin American Spanish.
Technologies used: Python, TensorFlow, Mongodb.
TEACHING EXPERIENCE
Private University
Sep 2018 – Nov 2018
Teacher
Taught about artificial intelligence and signal processing to students and Supervisor of graduation works
Technologies used: Python, Matlab, Google class room
Private University
Aug 2018 – Dec 2018
Research Professor
Researched about neural networks applied to civil engineering. Professor of Numerical Methods, Static and Dynamic, and elective writing. Also developed the web of the civil engineering program.
Technologies used: Python, Google class room, HTML,CSS,Java Script
COMPANY EXPERIENCE
AGM Consultores LTDA
Mar 2013 – Dec 2013
System and Development Analyst
Researched new technologies for speech recognition. Responsible for the development of software to process speech. Developed a speech enhancement system in Python for an integration with an automatic recognition unit. Recorded audio to create acoustic models. Built a Python character classifier to elaborate language models. In charge of the test approach, test scenario and test plan.
Technologies used: Python, Git, Praat
ABMTENC
Feb 2012 – Feb 2013
Technical Assistant and Software Support
Provided software and hardware maintenance. Diagnosis and correction of errors found during the use of the programs. Adapt the system to changes that may occur in hardware, operating system, peripherals and work tools
Technologies used: LabView, Matlab, OS LINUX
SENA
Jan 2010 – Jan 2011
Engineering Assistant
Provided software and hardware maintenance. Was adviser of mechatronic automation projects. Designed and implemented electronic telemetry systems. Controlled networks and communications systems
Technologies used: LabView, XBee, Bash(Unix Shell, Proteus
HONORS & AWARDS
Best Paper Award.
Amsterdam, Dec 2017.
World Academy of Science, Engeneering and Technology.
Conference: ICSRL: 19th International Conference on Speech Recognition and Language Processing.
Paper: Speech Enhancement Using Wavelet Coefficients Masking with Local Binary Pattern.
Ph.D Scholarship, Student Grade 10.
Brazil, Aug 2015.
Center for Telecommunications Studies (CETUC).
M.Sc Scholarship.
Brazil, Mar 2013.
Center for Telecommunications Studies (CETUC).
PORTFOLIO
HOME BANKING
Basic Home Banking application made in JS, where the user of the application can see the balance of his account and simulate actions such as withdrawal, deposit and money transfers or payment of services.
PUZZLE
A puzzle game made in JS, on a dynamic web page, with movement at the user's request.
PIXEL ART
A painting game made in JS, where you create a tool to make different drawings by painting pixel by pixel.
ZOMBIE CITY
2D interactive game made in Js, where the main player has to reach the goal avoiding different obstacles and avoiding being caught by zombies or the uncontrolled train.
BOOKING
Dynamic web application made in JS for restaurant reservations.
SURVEYS
Web application to create and manage surveys with multiple questions and their possible answers from an administrator interface.
WHAT DO I SEE?
Web application for movie recommendation and rating, using genre or title filters. Developed in MySQL, Node Js and Express.
MOVIE vs MOVIE
Web application for movies' poll creations and management, through a back-end development with Express Node JS and MySQL.
