With recent advancements in computational intelligence, deep learning has gained increased attention from many artificial intelligence (AI) researchers due to its applicability in various areas, including e-healthcare, autonomous cars, surveillance systems, or remote sensing, among others. Deep learning models have the ability to automatically extract the potential features of the given data and so do not require any kind of hand-crafted features for the model building process. However, deep learning models require high computational power and resources, therefore, these models are not well suited for lightweight devices such as mobiles and the Internet of Things (IoT). Additionally, these models require an efficient tuning of the hyper-parameters.

To overcome these problems, we must optimize the architecture and initial parameters of deep learning models in such a fashion that it can be implemented on light weight devices. However, the optimization of deep learning models is a challenging problem since it may compromise the performance. Therefore, light weight deep learning models should be developed in such a fashion that they take less resources for optimal architecture and at the same time improve performance. To achieve this, researchers have started utilizing metaheuristic techniques to efficiently select the initial parameters of deep learning models. Still, the optimization of deep learning architecture is necessary to further investigate the structural and functional properties of these models for lightweight devices.

This Special Issue will provide a platform for researchers to share cutting-edge solutions in the field and to promote research and development activities in light weight deep learning models for multimodal data by publishing high-quality original research and review articles in this rapidly growing interdisciplinary field.

Potential topics include but are not limited to the following:

• Lightweight deep learning models
• Metaheuristics-based deep learning models
• Hardware for lightweight deep learning models
• Explainable lightweight deep learning models
• Lightweight deep reinforcement learning models
• Lightweight deep generative adversarial models
• Lightweight explainable machine learning models
• Lightweight deep recurrent neural networks
• Lightweight deep transfer learning models
• Lightweight deep learning models for Internet of Things
• Lightweight deep learning models for medical devices