Put image recognition AI in real world gaming application.
An aimbot/aim assistant for the game Valorant using image recognition techniques. The program uses YOLOv5 Object Detection Algorithm to locate enemies on the screen and smooths the aiming path by adding stops and using randomization techniques. The program then feeds the smoothed aiming path to an external Arduino device, which acts as an external mouse and was not detectable by Vanguard anti-cheat software.
- YOLOv5
- Python
- Arduino
- Roboflow
- Successfully implemented image recognition skill to locate enemies on the screen
- Developed a smoothing algorithm to create a more randomise yet accurate aiming path
- Integrated the program with an external Arduino device for bypassing the world best anti-cheat detection
- Designed a functional and user-friendly IMGUI interface for the program
- Created a custom dataset Yolov5 that is publicly accessible
- The dataset was selected by Roboflow to be featured in the Roboflow universe
toflan.mp4
i7z9dw.mp4
tx7yin_1.mp4
https://universe.roboflow.com/aimbot-creators/valorant-object-detection2/dataset/4
I am highlighting proficiency in creating custom image reconition datasets and showcasing my work on a platform recognized by industry experts. This achievement also demonstrates my ability to work with open-source projects and contribute to the larger community.
Disclaimer This software is for educational purpose and may use this code to learn to build image recognition application.