Video Gallery
Collection of videos from Team Five AI Aiolearn in various AI projects and challenges
Our Project Videos
Web Challenge #1
Our solution for this challenge was to design an interactive web system that, upon receiving JSON data, simulates robots as rectangles of various sizes on a two‑dimensional grid. Using intelligent placement algorithms such as Best Fit, First Fit, and Bottom‑Left, the system determines robot positions without overlap, minimizes empty space, and visually displays the results along with the precise location of each robot.
Web Challenge #2
Our solution for this challenge was to develop a browser‑based intelligent traffic light control system that uses queue algorithms, prioritization, and dynamic decision‑making to optimize traffic flow at a four‑way intersection. The system detects the busiest paths in real time, selects the optimal direction to turn green, and manages vehicle passage based on priority.
Web Challenge #3
Our solution for this challenge was to implement a browser‑based interactive game simulating pathfinding on a square grid. The map is built from a predefined two‑dimensional matrix where values represent free paths, obstacles, start, and end points. The player clicks cells to define their path step by step, with the chosen path highlighted in real time. At the end, the system validates the route by checking passage through valid cells, reaching the end point from the start, avoiding repetitions, and not colliding with obstacles, then displays success or rejection accordingly.
Web Challenge #4
In this challenge, the task is to build a fully client‑side encryption‑decryption tool that transforms plain text into a unique, browser‑generated image, and reverses the process to recover the original text. The system uses an algorithm to assign each character from A–Z a distinct graphical pattern, such as a specific color combined with a fixed position or shape, all created through HTML elements and CSS rather than pre‑made images. When encrypting, the input text is converted into a grid of colored or shaped blocks, each visually representing a character according to the pattern rules. When decrypting, the tool reads the block positions, colors, or shapes, matches them to their corresponding characters, and reconstructs the text. All logic for both image generation and reading runs entirely in the browser, ensuring the process is self‑contained with no server or file dependencies.
AI Challenge #1
To support decision-making, we also created an interactive Streamlit dashboard to visualize high-risk areas, enabling authorities to take preventive actions proactively.
AI Challenge #2
An interactive AI‑powered virtual avatar platform that blends real‑time 3D rendering, and expressive animations with a personality‑driven chatbot. Built with Three.js, FastAPI, and OpenAI integration for engaging, lifelike conversations.
AI Challenge #3
In the next stage, the platform employs cutting-edge image generation models like DALL-E 3 from OpenAI to create artworks that not only resonate with the user’s emotional state but also leave a profound visual and emotional impact.
AI Challenge #4
We have built a comprehensive system utilizing YOLOv8 for object and person detection, MediaPipe Pose for body posture analysis, and custom Action Recognition algorithms.