2022 at Lockheed Martin
Machine Learning
- “SAD: Self-supervised Avionic Diagnostics”. In proceedings of SPIE 2023 Conference: Defense + Sensing (SPIE '23). Accepted with ORAL presentation. Maksim Bobrov, Attiano Purpura-Pontoniere, Tarun Bhattacharya
Who Am I?
Cogito ergo sum + Ubuntu - I think therefore I am, however I am because we are.
I am pedantic and semantic. I like homonyms, homophones, homographs, synecdoches, stenography, steganography (No I didn't hide any messages in the opaque Basquiats, trust me looking through the binary of those would be a waste of time, seriously), and other estoteric words.
I have an idiosyncratic philosophy that stems from a syncretic snythesis of eastern & western religions and science.
Like most engineers I am motivated by understanding things, however I understand life is about people, and I put them first. I evaluate things based on: 1. The People, 2. The Place, 3. The Thing.
Basically I am Attiano Purpura-Pontoniere.
I am currently employed as a senior AI research engineer at Lockheed Martin's Advanced Technology Center in Palo Alto.
My research interests lie primarily in computer vision and stochastic optimization, which I view machine learning as a subset of.
Ways to find me include:
For detailed information on my past
simply visit my Linkedin!
My History
Projects
Papers
Technical Roles
My Papers
2021 at Lockheed Martin
Machine Learning
- “Efficient ATR Using Contrastive Learning”. In proceedings of SPIE 2022 Conference: Defense + Sensing (SPIE '22). Accepted with ORAL presentation. Attiano Purpura-Pontoniere, Abdullah Imran, Tarun Bhattacharya
2021 at UCLA & Stanford
Machine Learning
- Paper in semi-supervised learning submitted and rejected from CHIL 2022, ISBI 2022, MedNeurIPS 2022... will be published on ArXiv
2021 at Lockheed Martin
Machine Learning
- Whitepaper in Advanced HTI processing submitted to NGA BAA, was not recommended for further development. E. Berkson, A. Purpura-Pontoniere
2020 at Robotire
Machine Learning
- Robotire purchased IP for Lugnut Identification Using State of the Art Regional and Fully Convolutional Neural Networks, A. Purpura-Pontoniere, D. Terzopolous
Most Recent Projects
Current | Medical data analysis | 2
UCLA Cardiovascular Data Science Lab
Working with Dylan Steinecke to research and develop machine learning techniques for medical data analysis. We are currently investigating ML vs. deterministic algorithms for ECG compression for mobile health.
Current | Mentorship | 3
UCLA Eta Kappa Nu Alumni Mentorship Program
Mentor 3 students from UCLA HKN (electrical enginering honor society - implies consistent strong academic performance) 1-on-1 for 1 hour weekly, through the almuni mentorship program. Particpated in 2021, and 2022.
December 2021 | Computer Vision (segmentation, classification, denoising, etc) | 4
Stanford RSL - AI/ML Medical Image Analysis
Worked with Abdullah Imran when he was a post-doc at Stanford RSL to research and develop self-supervised and semi-supervised machine learning techniques for medical image analysis
December, 2020 | Object Detection | 2
Lug nut centriod identification
Worked with Robotire to provide a state of the art computer vision service that performs object detection of lug nuts in captured images through optical sensing
December, 2020 | NLP, NLG | 3
Rap Lyric Generator
Lead a team of M.S. students to create a rap lyric generator using SOTA NLP at the time. Now ChatGPT can probably do it better :/
June, 2020 | Quantum Computing | 3
Quantum Computer Evaluation
Lead a team of M.S. students and one PhD to evaluate IBM & Regetti quantum computers and simulators Pyquil and Qiskit implementations of standard QC algos
December, 2019 | Neuro signal processing | 1
EMG Signal Decomposition
Given real EMG data obtained from electrodes implanted in monkeys, used Matlab for signal decomposition to find 4 motor neurons. Used FFT to bandpass relevant frequencies, threshold spikes (action potentials) based on RMS power, and then use K-means on the top PCA eigenvalues of the filtered signal to cluster Motor Unit Action Potentials. Advised by Dr. Wentai Liu
March, 2019 | Computer Vision | 2
Handwritten Character Recognition
Used deep learning to perform optical character recognition on two datasets of increasing difficulty as the capstone project for my B.S. in EE at UCLA in a two person team with Brian Tehrani.
March, 2019 | Computer Vision | 2
UCLA Digital Signal Processing Lab
Utilized a DSP specific embedded system (TI's LCDK) to perform a series of digital signal processing tasks. Record and modulate speech, polygon recognition in images using the Hough Transform, and vowel recognition using deep learning and features extracted from the PSD of the input audio
June, 2018 | Robotics | 3
Autonomous Robocar
Contributed to an autonomous battery powered robot-car that avoids obstacles and finds the shortest path in a grid using Dijkstra's shortest path
June, 2017 | Software Engineering | 1
Battleship Videogame in C++
Used Object Oriented Programming in C++ to make a battleship game that output text to the command line as a GUI. Created three levels of intelligence for deterministic AI: easy player, mediocre player, and good player
June, 2016 | Software Engineering | 1
Coding For All
In high school I designed a two hour workshop to introduce young people with special needs to computer programming. Presented at the National Down Syndrome Congress Convention in Arizona, invited back the following year. Presented multiple times at the Silicon Valley Down Syndrome Network in Palo Alto.
Education
I received my M.S. in Computer Science, conferred December 2020 from University of California Los Angeles
- Primary Focus: ARTIFICIAL INTELLIGENCE: COMPUTER VISION, NATURAL LANGUAGE PROCESSING
- Secondary Focus: QUANTUM COMPUTING
I received my B.S. degree from University of California Los Angeles Cum Laude, conferred March 2019. My undergraudate capstone culminated in OCR using RNNs and CNNs on a handwritten names dataset
Professional Experience
Senior AI Research Engineer April 2022 ~ present
AI Research Engineer April 2021 ~ April 2022
At Lockheed Martin, advised by Dr. Tarun Bhattacharya
- Technical lead (key technical contributor + PM responsibilities) for a team developing a high priority cognitive application.
Awards:
- Awarded 2 publications
- Delivered 9 talks on work & research - 2 external, 7 internal lightning talks
- Submitted 3 invention disclosures
- Proposed 3 IRADs
- Nominated for Exceptional Engineering and Technology Award 2022
- Winner of the Space Innovation Spotlight 2022
- Nominated for Space Awards Night 2022
- Featured on Space 2022 Greatest Hits
- Winner of Data Science Sponsored Innovation 2021
- Awarded NextGen Recognition from direct supervisors and senior engineers
- Promoted in 1st year
Computer Vision Engineer July 2020 ~ December 2020
At Robotire, advised by Mr. Bradley Sliz
- Specified, designed, and implemented custom object detection services from scratch. Integrated with industral articulated robots.
- Created an LLC to manage contract and liability. (PurpleBridgeMakerTech LLC)
Software Engineer May 2019 ~ July 2020
At Raytheon Space and Airborne Systems, advised by Mr. Omar Channer
- Injected random noise into SAR data→had a consensus of classifiers identify/correct mislabels resulting in >96% correction
- Built a variable-size data augmenter w/ complex rotation, Rayleigh and Gaussian noise, improved classification by >14%
- Developed a python script to automate the plotting of ROC curves given a classifier and a dataset using Tensorflow
- Exstensive low-level (ELF object code) debugging of instrumentation/BDB maker for EO sensor servo codebase in SVN branch, pushed code that fixed Jenkins build
Software Development Intern (UX/UI) June 2018 ~ September 2018
At HSQ Technologies, advised by Mr. Peter Polissky
- Specified, designed, and implemented custom object detection services from scratch. Integrated with industral articulated robots.
- Created an LLC to manage contract and liability. (PurpleBridgeMakerTech LLC)
Electrical Engineering Lab Mentor August 2017 ~ June 2018
At the UCLA EE department, advised by Dr. Dennis Briggs
- Coached groups of 3-5 undergraduate EE students through assigned lab projects such as verifying Kirchhoff's laws
- Provided hands-on guidance to groups as they built a line-following robocar using motors, phototransistors and arduino nanos
Lead Tutor August 2012 ~ June 2016
At Peralta Community Colleges, advised by Prof. Pinar Alscher and Prof. Jose Luis
- Tutored students one-on-one or small groups in subjects such as Math, Chemistry, Physics and Computer Science. Hand-selected by Chair of Computer Science department at Laney College (Jose Luis Flores) to tutor student with autism, to positive feedback.
I was a teaching assistant at:
- School Year 2020 CS M51A: Logic of Digital System Design under Professor Richard Korf
- School Year 2020 CS M151B: Software Architecture under Professor Glenn Reinman
- School Year 2020 CS 33: Computer System Organization under Professor Glenn Reinman
- School Year 2020 (summer quarter) CS 33: Computer System Organization under Professor Parvenh Ghaforyafard
Oh, and I almost forgot, I am on a Grammy award winning choral performance of Mahler's 8th symphony.