Submission of Work on 2025 CUES Hackathon from Group 14 #4
JasperWANG-911
started this conversation in
Ideas
Replies: 1 comment
-
Thank you guys for sharing your work and apologies for the delay. I will link your work to the page shortly |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
Project Overview
Our project was inspired by recent research on dynamic AI agents for interpreting remote sensing images, which we sought to apply in the education sector. Teachers often struggle to track students’ performance and engagement in real-time, so we aimed to develop an AI-powered solution to enhance teaching efficiency. GitHub Repo: https://github.com/JasperWANG-911/AI_Agent
How It Works
The system utilizes a computer’s webcam to continuously capture images of the classroom, storing them for later analysis. Teachers can then input queries into our website, and our dynamic AI agent will identify relevant images, student exam records, and reports. Using this information, the agent generates a comprehensive report in response to the query, helping overworked teachers quickly assess student performance with minimal effort.
Most distinctive feature
Our most distinctive feature is our dynamic AI agents. Unlike traditional static agents that perform fixed tasks, our dynamic agents adapt their strategies using built-in intelligence and can integrate a variety of pre-defined functions(called APIs) as needed. This approach offers two key advantages:
Enhanced Accuracy and Flexibility: To boost the accuracy of predictions and analyses, you simply add more pre-defined functions tailored to your needs. Over time, as we continuously expand this function pool, the agents will be capable of handling an increasingly diverse range of queries.
Improved Efficiency and Cost Savings: This dynamic method is not only smarter but also more efficient in terms of algorithm usage and cost. For example, consider a scenario where our API pool grows to include thousands of functions, yet you ask a simple question like, "How did Lisa perform in her math exam recently?" In this case, the agent will automatically bypass complex processes which are not helpful at all, thereby saving significant computational resources and reducing costs.
Contributors
Feel free to reach out to us if you find this dynamic agent interesting!
Beta Was this translation helpful? Give feedback.
All reactions