RenderGAN: Generating Realistic Labeled Data
This addresses the data annotation bottleneck for computer vision researchers, though it is incremental as it builds on existing GAN and 3D modeling techniques.
The paper tackles the problem of high annotation costs for training deep convolutional neural networks by introducing RenderGAN, a framework that generates realistic labeled images using a 3D model and GANs, and shows that training a DCNN on this generated data yields considerably better performance than baselines.
Deep Convolutional Neuronal Networks (DCNNs) are showing remarkable performance on many computer vision tasks. Due to their large parameter space, they require many labeled samples when trained in a supervised setting. The costs of annotating data manually can render the use of DCNNs infeasible. We present a novel framework called RenderGAN that can generate large amounts of realistic, labeled images by combining a 3D model and the Generative Adversarial Network framework. In our approach, image augmentations (e.g. lighting, background, and detail) are learned from unlabeled data such that the generated images are strikingly realistic while preserving the labels known from the 3D model. We apply the RenderGAN framework to generate images of barcode-like markers that are attached to honeybees. Training a DCNN on data generated by the RenderGAN yields considerably better performance than training it on various baselines.