• Vikas Kumar
• Ayan Rajput

Efficient and precise car recognition is essential in the age of driver-assistance systems (ADAS) and driverless vehicles. The "You Only Look Once" (YOLO) method, a cutting-edge real-time object detection model renowned for its speed and accuracy, is used in this paper's automobile object detection system. YOLO treats object detection as a single regression issue, taking bounding box coordinates and class probabilities directly from image pixels.We trained the YOLO model on a diverse dataset comprising images of cars in various environments, lighting conditions, and occlusions to enhance the system's robustness. The dataset includes both urban and highway scenes, providing a comprehensive benchmark for evaluating the model's performance. Our implementation leverages YOLOv4, an improved version of the original YOLO, which incorporates several architectural enhancements such as CSPDarknet53 as the backbone, a Path Aggregation Network (PANet) for feature aggregation, and a spatial pyramid pooling layer to increase the receptive field. These improvements contribute to better object localization and classification accuracy.The results demonstrate that our YOLO-based car detection system achieves high precision and recall rates, with a mean average precision (mAP) exceeding 80% across our test dataset.

Keywords: YOLO, Object Detection, Car Detection, Autonomous Vehicles, Real-Time Detection, Computer Vision.

"CAR OBJECT DETECTION USING YOLO", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 7, page no.g110-g117, July-2024, Available :http://www.jetir.org/papers/JETIR2407613.pdf

"CAR OBJECT DETECTION USING YOLO", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 7, page no. ppg110-g117, July-2024, Available at : http://www.jetir.org/papers/JETIR2407613.pdf