A Deep Learning Approach for Vehicle Detection
Authors:
- 1 December 2018
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 98-102
Abstract
The autonomous driving needs some several features to achieve driving without human interference. One of these features is vehicle classification and detection since the target of this process is to help the CPU ''Central Processing Unit" of the vehicle to see what is around the vehicle, in order to evaluate the situation to take the best decision for each situation in real time. This paper is focusing on the classification process of the video-based vehicle detection, to achieve that, different deep learning techniques have been implemented which are known as convolutional neural networks (CNN) architectures. These CNN architectures are ResNet, Inception-ResnetV2, InceptionV3, NASNet, MobileNetV2, and PNASNet architectures. Also there are two different datasets have been trained in these architectures to evaluate them. These datasets are Kitti dataset to train on car detection only, in additions to MIO-TCD dataset to detect various types of vehicles. The Inception-ResnetV2 have shown the best performance in our results.Keywords
This publication has 10 references indexed in Scilit:
- ImageNet classification with deep convolutional neural networksCommunications of the ACM, 2017
- SSD: Single Shot MultiBox Detector2016
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal NetworksIEEE Transactions on Pattern Analysis and Machine Intelligence, 2016
- Deep Residual Learning for Image Recognition2016
- Rethinking the Inception Architecture for Computer Vision2016
- Caffe2014
- Deep, Big, Simple Neural Nets for Handwritten Digit RecognitionNeural Computation, 2010
- Object Recognition with Features Inspired by Visual Cortex2005
- Comparison of classifier methods: a case study in handwritten digit recognition2002
- I.—COMPUTING MACHINERY AND INTELLIGENCEMind, 1950