Coupling Rendering and Generative Adversarial Networks for Artificial SAS Image Generation
This addresses data scarcity for SAS imaging in marine applications, but it is incremental as it builds on existing GAN and rendering techniques.
The paper tackles the problem of limited and skewed Synthetic Aperture Sonar (SAS) datasets by introducing SAS GAN, a pipeline that couples an optical renderer with a GAN to generate realistic SAS images of seafloor targets, showing potential for more effective dataset augmentation than an off-the-shelf GAN.
Acquisition of Synthetic Aperture Sonar (SAS) datasets is bottlenecked by the costly deployment of SAS imaging systems, and even when data acquisition is possible,the data is often skewed towards containing barren seafloor rather than objects of interest. We present a novel pipeline, called SAS GAN, which couples an optical renderer with a generative adversarial network (GAN) to synthesize realistic SAS images of targets on the seafloor. This coupling enables high levels of SAS image realism while enabling control over image geometry and parameters. We demonstrate qualitative results by presenting examples of images created with our pipeline. We also present quantitative results through the use of t-SNE and the Fréchet Inception Distance to argue that our generated SAS imagery potentially augments SAS datasets more effectively than an off-the-shelf GAN.