Real Time Human and Object Detection with Yol...

Real Time Human and Object Detection with Yolov3

Author Name : Chetlapelly Sai Kiran Goud, I. Govardhana Rao

DOI: https://doi.org/10.56025/IJARESM.2023.1182365

ABSTRACT Real-time live human detection is a critical task in computer vision and deep-learning techniques with numerous applications, such as security, surveillance, and robotics. In recent years, deep learning-based object detection methods have demonstrated remarkable in accurately detecting objects in real-world scenarios. Among these methods, the YOLOv3(You Only Look Once version3) algorithm emerged as a popular algorithm for its outstanding speed and competitive accuracy. It is single-short algorithm, which means that it predicts bounding boxes and class probabilities for all objects in a image in single pass. This makes YOLOv3 very fast, making it suitable for real time applications. In this paper, we propose a method for a real time live human detection based on YOLOv3. We leverage the power of convolutional neural networks(CNNs) and YOLOv3’s efficient architecture to achieve near real-time performance, ensuring rapid and reliable human detection in dynamic environments. Our method uses pre-trained YOLO3 model to detects human in live video feed. We evaluate real time performance while maintaining high accuracy. Our results show that our method is a promising approach for real time live human detection method is fast, accurate and scalable.