Deep learning has made various breakthroughs in cognition tasks. However, deep learning neural networks, by nature, are not only computation-intensive but also memory-intensive. Therefore, most of the existing deep learning applications are cloud-based, which then raises the concern of latency, efficiency, and bandwidth consumption. It is critical to tune the neural network models to accelerate the execution of deep learning applications on embedded systems. Even with the breakthroughs in the FPGAs or ASICs tailored for deep learning, it is proven necessary to enhance the neural network models using modern computational optimization approaches for more efficient execution
Инноватика-2021 : сборник материалов XVII Международной школы-конференции студентов, аспирантов и молодых ученых, 22-23 апреля 2021 г., г. Томск, Россия. Томск, 2021. С. 106-108