A survey of the recent architectures of deep convolutional neural networks
- PDF / 3,137,553 Bytes
- 62 Pages / 439.37 x 666.142 pts Page_size
- 50 Downloads / 176 Views
A survey of the recent architectures of deep convolutional neural networks Asifullah Khan1,2 · Anabia Sohail1,2 · Umme Zahoora1 · Aqsa Saeed Qureshi1
© Springer Nature B.V. 2020
Abstract Deep Convolutional Neural Network (CNN) is a special type of Neural Networks, which has shown exemplary performance on several competitions related to Computer Vision and Image Processing. Some of the exciting application areas of CNN include Image Classification and Segmentation, Object Detection, Video Processing, Natural Language Processing, and Speech Recognition. The powerful learning ability of deep CNN is primarily due to the use of multiple feature extraction stages that can automatically learn representations from the data. The availability of a large amount of data and improvement in the hardware technology has accelerated the research in CNNs, and recently interesting deep CNN architectures have been reported. Several inspiring ideas to bring advancements in CNNs have been explored, such as the use of different activation and loss functions, parameter optimization, regularization, and architectural innovations. However, the significant improvement in the representational capacity of the deep CNN is achieved through architectural innovations. Notably, the ideas of exploiting spatial and channel information, depth and width of architecture, and multi-path information processing have gained substantial attention. Similarly, the idea of using a block of layers as a structural unit is also gaining popularity. This survey thus focuses on the intrinsic taxonomy present in the recently reported deep CNN architectures and, consequently, classifies the recent innovations in CNN architectures into seven different categories. These seven categories are based on spatial exploitation, depth, multi-path, width, feature-map exploitation, channel boosting, and attention. Additionally, the elementary understanding of CNN components, current challenges, and applications of CNN are also provided. Keywords Deep learning · Convolutional neural networks · Taxonomy · Representational capacity · Residual learning · Channel boosted CNN
* Asifullah Khan [email protected] 1
Pattern Recognition Lab, DCIS, PIEAS, Nilore, Islamabad 45650, Pakistan
2
Deep Learning Lab, Center for Mathematical Sciences, PIEAS, Nilore, Islamabad 45650, Pakistan
13
Vol.:(0123456789)
A. Khan et al.
1 Introduction Machine Learning (ML) algorithms are known to learn the underlying relationship in data and thus make decisions without requiring explicit instructions. In literature, various exciting works have been reported to understand and/or emulate the human sensory responses such as speech and vision (Hubel and Wiesel 1962, 1968; Ojala et al. 1996; Chapelle 1998; Lowe 1999; Dalal and Triggs 2004; Bay et al. 2008; Heikkilä et al. 2009). In 1989, a new class of Neural Networks (NN), called Convolutional Neural Network (CNN) (LeCun et al. 1989) was reported, which has shown enormous potential in Machine Vision (MV) related tasks. CNNs are one of the best lea
Data Loading...
