Paper Title
Heterogeneous Hand Gesture Detection Using Yolo-V4 Model

Abstract
Training custom dataset with YOLO to attain faster performance and better accuracy is most desired modelnowadays to continue the progress in object detectionin the field of Artificial Intelligence. With this, detection performance can be optimized directly as YOLO performs on full images. Expressing signs trough hand strokes is one of the fundamental strategies of human communication. So many contributions have been made in this field over the past years, yet this research attempts to build a faster model using very little dataset compared to plethora of data to fed to the model which ultimately requires longer time to train the model and lacks in accuracy as well. Eloquent and mostly used Indian subcontinental hand gestures have been used as dataset, involving single and both hand in order to develop a better man-machine communication. Ten different types of datasets have been trained, validated and tested trough YOLO-V4 Model after labelling thoroughly and compared to other existing model for better justification. Keywords - YOLO, YOLO-V4, Hand Gesture, Hand Gesture Recognition, Static Hand Gesture.