With the rapid developments of artificial intelligence (AI) and deep learning (DL) techniques,
it is critical to ensure the security and robustness of the deployed algorithms. Recently, the …

It is commonly believed that networks cannot be both accurate and robust, that gaining
robustness means losing accuracy. It is also generally believed that, unless making …

Edge intelligence has played an important role in constructing smart cities, but the
vulnerability of edge nodes to adversarial attacks becomes an urgent problem. A so-called …

In adversarial machine learning, attackers add carefully crafted perturbations to input, where
the perturbations are almost imperceptible to humans, but can cause models to make wrong …

近年来, 以机器学习算法为代表的人工智能技术在计算机视觉, 自然语言处理,
语音识别等领域取得了广泛的应用, 各式各样的机器学习模型为人们的生活带来了巨大的便利 …

Dynamic convolution achieves better performance for efficient CNNs at the cost of negligible
FLOPs increase. However, the performance increase can not match the significantly …

The transformer architecture has revolutionized bioinformatics and driven progress in the
understanding and prediction of the properties of biomolecules. Almost all research on large …

Quantizing neural networks to low-bitwidth is important for model deployment on resource-
limited edge hardware. Although a quantized network has a smaller model size and memory …

The advent of deep learning has revolutionized the technology industry and has made Deep
Neural Networks (DNNs) the powerhouse of many modern day software applications. Well …

This dissertation tackles the issues of improving and evaluating the robustness of machine
learning models for medical diagnosis. We describe a data augmentation technique that …