A Comprehensive Summary of Three Years of Research Contributions in Computer Vision, Machine Learning, and Beyond (2019-2022)
We have made contributions in the fields of computer vision, pattern recognition, machine learning, artificial intelligence, computation and language, cryptography and security, and software engineering. Our works have focused on various topics such as rapid object detection, sub-typing, text augmentation, image compression, super resolution, and adversarial risk to machine learning models. Our research has also included automating defense against adversarial attacks and discovering model characteristics through strategic probing. Additionally, we have applied our expertise to the gaming world by developing generative language models for chess and go gameplay. Our papers have been published in various formats and conferences, including SPIE, AI4I, NIAI, ICAIT, and NFCS. This is what we did.
Computer Vision and Pattern Recognition, Machine Learning, Artificial Intelligence, Computation and Language, Image and Video Processing, Cryptography and Security, Software Engineering, Adversarial Risk, Adversarial Attacks, Neglected Languages, Generative Language Models, Game Play, Strategic Probing, White Box Model Characteristics
The following work represents a team's efforts on various research questions. I will mention the other authors here and recognize their efforts: David Noever, Josh Kalin, Dominick Hambrick, Willie Maddox, Grant Rosario and Wes Regian. Thank you all for your intellectual conversations and commentary over the years on these papers and on the Machine Learning industry as a whole. The following paper represents a subset of our efforts over the past three years to the following subjects
| Subject | Subject Count |
|---|---|
| Machine Learning (cs.LG) | 10 |
| Computer Vision and Pattern Recognition (cs.CV) | 7 |
| Computation and Language (cs.CL) | 5 |
| Cryptography and Security (cs.CR) | 4 |
| Artificial Intelligence (cs.AI) | 2 |
| Image and Video Processing (eess.IV) | 2 |
| Software Engineering (cs.SE) | 1 |
| Computer Science and Game Theory (cs.GT) | 1 |
| TOTAL PAPERS | 16 |
A summary of all the papers can be viewed in the Research Overview file. Individual papers are also in this repository and are listed in chronological order below as seen on arXiv.
| Paper | Year | Month | Authors | File |
|---|---|---|---|---|
| Soft Labels for Rapid Satellite Object Detection | 22 | 12 | Matthew Ciolino, Grant Rosario, David Noever | |
| Soft-labeling Strategies for Rapid Sub-Typing | 22 | 09 | Grant Rosario, David Noever, Matt Ciolino | |
| The Turing Deception | 22 | 12 | David Noever, Matt Ciolino | |
| Back Translation Survey for Improving Text Augmentation | 21 | 02 | Matthew Ciolino, David Noever, Josh Kalin | |
| Image Compression and Actionable Intelligence With Deep Neural Networks | 22 | 03 | Matthew Ciolino | |
| Enhancing Satellite Imagery using Deep Learning for the Sensor To Shooter Timeline | 22 | 03 | Matthew Ciolino, Dominick Hambrick, David Noever | |
| Color Teams for Machine Learning Development | 21 | 10 | Josh Kalin, David Noever, Matthew Ciolino | |
| A Modified Drake Equation for Assessing Adversarial Risk to Machine Learning Models | 21 | 03 | Josh Kalin, David Noever, Matthew Ciolino | |
| Fortify Machine Learning Production Systems: Detect and Classify Adversarial Attacks | 21 | 02 | Matthew Ciolino, Josh Kalin, David Noever | |
| Automating Defense Against Adversarial Attacks: Discovery of Vulnerabilities and Application of Multi-INT Imagery to Protect Deployed Models | 21 | 03 | Josh Kalin, David Noever, Matthew Ciolino, Dominick Hambrick, Gerry Dozier | |
| Local Translation Services for Neglected Languages | 21 | 01 | David Noever, Josh Kalin, Matt Ciolino, Dom Hambrick, Gerry Dozier | |
| The Chess Transformer: Mastering Play using Generative Language Models | 20 | 08 | David Noever, Matt Ciolino, Josh Kalin | |
| The Go Transformer: Natural Language Modeling for Game Play | 20 | 07 | Matthew Ciolino, David Noever, Josh Kalin | |
| Black Box to White Box: Discover Model Characteristics Based on Strategic Probing | 20 | 09 | Josh Kalin, Matthew Ciolino, David Noever, Gerry Dozier | |
| Training Set Effect on Super Resolution for Automated Target Recognition | 19 | 11 | Matthew Ciolino, David Noever, Josh Kalin | |
| Discoverability in Satellite Imagery: A Good Sentence is Worth a Thousand Pictures | 20 | 01 | David Noever, Wes Regian, Matt Ciolino, Josh Kalin, Dom Hambrick, Kaye Blankenship |