INTEGRATING YOLOV8 AND GOOGLE MEDIAPIPE FOR ADVANCED PEDESTRIAN DIRECTION ANALYSIS IN AUTONOMOUS VEHICLES
In this project, we are looking into combining YOLOv8 and Media Pipe to enhance pedestrian detection in autonomous vehicles. Our goal is to increase the accuracy and efficiency of recognizing and estimating pedestrians' poses in city settings by using pre-trained YOLOv8M model trained on the COCO dataset.
As self-driving cars become more common in city streets, it's vital to have reliable pedestrian detection to keep everyone safe. This effort focuses on solving this important problem by combining the capabilities of two cutting-edge technologies: YOLOv8 for accurate detection and Media Pipe for precise pose estimation.
- Pedestrian Detection Accuracy with YOLOv8
- Contribution of Media Pipe to Autonomous Vehicle Perception
- Integration of YOLOv8 and Google Media Pipe
- Set Up a Python Virtual Environment
- Install Required Libraries from requirements.txt
- once all the libraries are installed, add images into the input_directory.
- run the img.py file
- the output will be saved in output_img directory.
Short-Term Goals:
- Improve handling of occlusions in pedestrian detection.
- Enhance the model's robustness to different lighting conditions.
Long-Term Goals:
- Explore real-time video processing capabilities for dynamic pedestrian detection.
- Develop a system that autonomously adapts to new urban environments and pedestrian behaviors.