An Overview of Recent Convolutional Neural Network Algorithms for Image Recognition
Genevieve Sapijaszko, Wasfy B. Mikhael · 2018
An important aspect of a guidance system in an autonomous vehicle is the detection of objects and extraction of reliable features that can identify the object from images. Object detection methods have evolved in the last 20 years, with convolutional neural network algorithms in particular showing promise. This paper compares and contrasts recent convolutional neural network algorithms using the BelgiumTS Dataset. The different convolutional neural networks used and compared are AlexNet, VGGNet, GoogleNet, and ResNet. The object detection methods will be used to find the best processing speed and recognition rate among the convolutional neural networks. A survey of published state-of-the-art, convolutional neural networks are evaluated against published results.