Deep learning, as an important branch of artificial intelligence, has made significant progress in the fields of visual computing and signal processing in recent years. This study systematically explores the application of deep learning techniques in these two fields and reveals the powerful ability of deep learning in processing complex data and achieving high-precision tasks by analysing its core technologies, commonly used models and real-world application cases. In the field of visual computing, Convolutional Neural Networks (CNNs) are widely used in tasks such as image classification, target detection, image segmentation, and image generation and enhancement, which significantly improves the performance and application scope of computer vision. In the field of signal processing, models such as Recurrent Neural Networks (RNN) and Transformer have performed well in speech recognition, natural language processing, and biomedical signal analysis, driving the development of technologies such as intelligent voice assistants, real-time translation, and medical diagnosis. Nevertheless, deep learning still faces many challenges in terms of computational resource requirements, data annotation and acquisition, model interpretability and transparency. By analysing these issues in detail, this paper proposes directions for future research, aiming to provide valuable references and guidance for academia and industry.