Laboratory of Applications in Informatics and Analytics
Background Information
Learn how computer vision works (image recognition and object detection models).
Create a mobile app that runs an object detection model on user-inputted data.
Explore various machine learning and software development platforms.
Learn about the fields of Artificial Intelligence and Computer Science.
Team
Mahdi Belcaid
Internship Mentor
Keora Flanary-Olayvar
Academic Mentor
Taryn Takebayashi
Academic Mentor
Kieran Lynch
Student Intern
Raquel Fujita
Student Intern
Sasha Watanabe
Student Intern
Tyler Ahina
Student Intern
Student Intern Presentations and Reflections
Kieran Lynch
This internship taught a lot about coding, but also no-code app development. In addition with training and teaching machines to detect and classify objects, we spent days learning how to convert those machines into a tool usable by people.
Learning was a quick and simple process, and implementing our ideas as a group, rather than individuals made our final projects better than they would have been otherwise.
Raquel Fujita
Before combining our knowledge during the final AppGyver project, we independently experimented with Plainsight and Teachable Machine to better understand how to train an image classification model. Within Teachable Machine, we trained our models to correctly categorize images during tests—I trained my model to recognize different Marvel characters! Before working with Teachable Machine, through lecture-style sessions, we were taught how to define and apply machine learning regression, classification, and dataset terms, such as underfitting and overfitting. Afterward, through Replit, we all practiced with for loops, while loops, and functions, to name a few facets of Python programming, that provided detailed insight into the larger-scale programming of mature classification training models. We’d later recognize those Python applications when working with code in Jupyter Notebook. Utilizing the Replit program allowed us to perform tasks by building on simpler code. Consequently, this internship provided me with a solid foundation of artificial intelligence and object detection knowledge that I would enjoy applying to other disciplines or projects.
Throughout the extent of the internship, we progressed from learning the foundation of image classification and artificial intelligence through Teachable Machine to practicing with Python code used to train such classification models. I’d like to apply similar concepts under the branch of Data Analytics in Civil and Transportation Engineering practices I plan to study during college. I’d be interested in learning more about objection detection approaches besides the two that we had practice with—using cropping tools and manual labeling. If given more time to experiment with AppGyver and Python, I would enjoy further individual practice with various Python tasks related to the no-code model. As Artificial Intelligence becomes increasingly prevalent in Transportation, I hope to use the knowledge gained during this internship and in the future to better my community.
Sasha Watanabe
Through a variety of hands-on activities, I learned the basics of how to train a computer model and use it in an app.
Even though my assignments weren't college-level work (thank God), it was interesting to learn from Professor Mahdi and Taryn. I also gained a lot of insight into what college is like--both at the undergrad and graduate levels--by talking with Taryn and Keora.
Tyler Ahina
In this internship, I learned a lot of things about coding and AI. I learned how the code communicates to the AI and how the AI communicates with the code.
When learning about code and AI was different from other things I have learned but it was pretty easy to understand.