Quantum Learning Transfer Model for the classification of people with safety helmets (#738)

Authors

Sulla-Torres, Jose

Martínez Muñoz, Jorge

Castillo Cáceres, César

Pacheco Oviedo, Abraham

Abstract

The problem of personal accidents in a work environment in many cases is due to the lack of use of helmets, among other safety elements. The objective of this work was to classify images of people who use or not use safety helmets, in the classes identified as correct use of the helmet, incorrect use and without a safety helmet, with a Quantum Transfer Learning model. For this, the method used has allowed the collection of a data set of 750 images of people in different work activities. The classical learning transfer model chosen was ResNet-18; the variational layers of the proposed model were built with the Basic Entangler Layers template for four qubits with the Pennylane quantum simulator. The results were 95.47% accurate in classifying the correct Safety Helmets. The conclusion reached is that the proposed model is a good option to detect and classify people with the use of safety helmets in a work environment.