Image synthesis with neural networks for traffic sign classification

Компьютерная оптика

Anton Konushin¹  Vladislav Shakhuro² 

[1] NRU Higher School of Economics, Moscow, Russia, Lomonosov Moscow State University, Moscow, Russia;NRU Higher School of Economics, Moscow, Russia;

关键词: traffic sign classification;    synthetic training sample;    generative neural network;   

DOI  :  10.18287/2412-6179-2018-42-1-105-112

来源: DOAJ

【 摘 要 】

In this work, we research the applicability of generative adversarial neural networks for generating training samples for a traffic sign classification task. We consider generative neural networks trained using the Wasserstein metric. As a baseline method for comparison, we take image generation based on traffic sign icons. Experimental evaluation of the classifiers based on convolutional neural networks is conducted on real data, two types of synthetic data, and a combination of real and synthetic data. The experiments show that modern generative neural networks are capable of generating realistic training samples for traffic sign classification that outperform methods for generating images with icons, but are still slightly worse than real images for classifier training.

【 授权许可】

Unknown