The printed circuit board (PCB) plays a critical role in any electronic product, and its manufacturing quality assurance is responsible for substantially impacting the final product’s price. Hence, research and development (R&D) for better inspection methods have been an internationally growing subject. Regarding the multiple PCB inspection paradigms, the visual investigation of these boards represents the majority of modern techniques. Thus, the present survey collects the most impacting studies over the last 25 years, highlighting their operation strategies and how they evolved until recently, including introducing artificial intelligence (AI) approaches to increase their overall performance and lower operating costs. Furthermore, this work calls attention to the importance of public PCB datasets to contribute to developing such methods and provide benchmarking references. Finally, the most recent challenges in the field are described and discussed.

This study introduces a sophisticated neural network structure for segmenting breast tumors. It achieves this by combining a pretrained Vision Transformer (ViT) model with a UNet framework. The UNet architecture, commonly employed for biomedical image segmentation, is further enhanced with depthwise separable convolutional blocks to decrease computational complexity and parameter count, resulting in better efficiency and less overfitting. The ViT, renowned for its robust feature extraction capabilities utilizing self-attention processes, efficiently captures the overall context within images, surpassing the performance of conventional convolutional networks. By using a pretrained ViT as the encoder in our UNet model, we take advantage of its extensive feature representations acquired from extensive datasets, resulting in a major enhancement in the model’s ability to generalize and train efficiently. The suggested model has exceptional performance in segmenting breast cancers from medical images, highlighting the advantages of integrating transformer-based encoders with efficient UNet topologies. This hybrid methodology emphasizes the capabilities of transformers in the field of medical image processing and establishes a new standard for accuracy and efficiency in activities related to tumor segmentation.