In semiconductor manufacturing, the wafer dicing process is central yet vulnerable to defects that significantly impair yield - the proportion of defect-free chips. Deep neural networks are the current state of the art in (semi-)automated visual inspection. However, they are notoriously known to require a particularly large amount of data for model training. To address these challenges, we explore the application of generative adversarial networks (GAN) for image data augmentation and classification of semiconductor wafer dicing induced defects to enhance the variety and balance of training data for visual inspection systems. With this approach, synthetic yet realistic images are generated that mimic real-world dicing defects. We employ three different GAN variants for high-resolution image synthesis: Deep Convolutional GAN (DCGAN), CycleGAN, and StyleGAN3. Our work-in-progress results demonstrate that improved classification accuracies can be obtained, showing an average improvement of up to 23.1 % from 65.1 % (baseline experiment) to 88.2 % (DCGAN experiment) in balanced accuracy, which may enable yield optimization in production.

The popular physics-based puzzle game series Angry Birds has been played and enjoyed by millions of people since its original launch in 2009. However, while the game may seem somewhat simple and straightforward to play, with even very young children being able to quickly grasp its mechanics and strategies, artificial intelligence has so far failed to obtain human-level performance. Along with a lack of knowledge about the game's internal physics engine and imprecise object detection algorithms, one of the core challenges to training better game-playing agents is the limited number and variety of available game levels. The levels in Angry Birds often contain individual structures that are made up of multiple rectangular 2D blocks, such as those shown in figure 1. While a handful of previous structure generators for Angry Birds exist, they often rely on hard-coded design constraints that limit the output diversity.

This paper investigates the suitability of using Generative Adversarial Networks (GANs) to generate stable structures for the physics-based puzzle game Angry Birds. While previous applications of GANs for level generation have been mostly limited to tile-based representations, this paper explores their suitability for creating stable structures made from multiple smaller blocks. This includes a detailed encoding/decoding process for converting between Angry Birds level descriptions and a suitable grid-based representation, as well as utilizing state-of-the-art GAN architectures and training methods to produce new structure designs. Our results show that GANs can be successfully applied to generate a varied range of complex and stable Angry Birds structures.