A Three-Player GAN: Generating Hard Samples To Improve Classification Networks
| Autor: | Davy Neven, Bert De Brabandere, Simon Vandenhende, Luc Van Gool |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
FOS: Computer and information sciences
Technology Discriminator Computer science Computer Vision and Pattern Recognition (cs.CV) Computer Science - Computer Vision and Pattern Recognition 02 engineering and technology Computer Science Artificial Intelligence 03 medical and health sciences 0302 clinical medicine Engineering Computer Science Theory & Methods 0202 electrical engineering electronic engineering information engineering Traffic sign recognition Science & Technology business.industry Pattern recognition Engineering Electrical & Electronic ComputingMethodologies_PATTERNRECOGNITION Computer Science 030221 ophthalmology & optometry 020201 artificial intelligence & image processing Artificial intelligence business Classifier (UML) Generative adversarial network |
| Zdroj: | MVA |
| DOI: | 10.48550/arxiv.1903.03496 |
| Popis: | We propose a Three-Player Generative Adversarial Network to improve classification networks. In addition to the game played between the discriminator and generator, a competition is introduced between the generator and the classifier. The generator's objective is to synthesize samples that are both realistic and hard to label for the classifier. Even though we make no assumptions on the type of augmentations to learn, we find that the model is able to synthesize realistically looking examples that are hard for the classification model. Furthermore, the classifier becomes more robust when trained on these difficult samples. The method is evaluated on a public dataset for traffic sign recognition. Comment: Accepted for oral presentation at MVA2019 |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|