The complexity of design rules and intense time-to-market demands have made auto-placement tools essential for advanced printed circuit board (PCB) designs. This paper presents a novel PCB placement framework to handle pad-to-pad clearance constraints and heterogeneous components to address these challenges. Unlike existing academic placers, our framework focuses on the following key features: a wire-area model to account for various routing resource needs between power and signal nets, a pad-to-pad clearance model to minimize spacing violations, and a two-sided, pad-type-aware density model to reduce component and pad overlap. We further develop a quadratic programming-based legalizer to resolve constraint violations among components of varying shapes. Experimental results show the effectiveness and efficiency of our framework, surpassing two state-of-theart academic placers in post-routing quality on both academic and industrial benchmarks.

Keyword spotting (KWS) system is an important human-computer interaction medium in smart devices. However, requiring the KWS model to maintain robust performance with a small number of parameters is extremely challenging. In addition to model architecture research, efficiently leveraging speech data to develop more powerful KWS systems is also an attractive research topic. In the speech-related community, various data augmentation methods have been studied. Mainstream methods mainly focus on manipulating speech waveforms or acoustic features. Regardless of the approach taken, a simple conclusion is that data augmentation strategies can further improve task performance. Motivated by practical needs and following interesting research lines, on the one hand, the paper presents an efficient KWS model that is designed to leverage the spatial and channel information of the input speech simultaneously while keeping the lightweight property of the model. On the other hand, we develop two novel data augmentation methods for training the KWS model to boost performance further. A series of experiments are conducted using the Google Speech Commands V2 dataset to evaluate the proposed KWS framework. Based on the experiments, the proposed framework can achieve 98.81% accuracy with fewer parameters than other state-of-the-art models.

Rapid breakthroughs in information technologies have driven substantial developments in artificial intelligence applications, particularly the widespread use of deep learning techniques in domains such as speech, image and text recognition. However, real world data distribution applications suffer from significant problems including data imbalance which can easily lead to machine learning biased towards the side with more data, resulting in inaccurate classification or prediction results. Therefore, effectively addressing data imbalance is a pressing research topic. Generative Adversarial Networks (GAN) addresses data imbalance, but is prone to vanishing gradients. Recent work has thus focused on improving the GAN architecture to resolve this problem. The present research extends these efforts, applying C4.5, Random Forest, Support Vector Machine, K-Nearest Neighbor and Naïve Bayes classification algorithms to a single imbalanced traffic collision dataset to identify methods for improving prediction results. Experimental results show that classification performance significantly improves after data augmentation using Synthetic Minority Oversampling Technique, GAN, Conditional GAN, and Gaussian Discriminant Analysis GAN as compared with the non-augmented dataset. In addition, the Gaussian Discriminant Analysis GAN with Naïve Bayes classifier produces a dataset that optimizes classification performance for traffic accident prediction at highway intersections.