Contemporary hardware implementations of artificial neural networks face the burden of excess area requirement due to resource-intensive elements such as multiplier and non-linear activation functions. The present work addresses this challenge by proposing a resource-efficient Coordinate Rotation Digital Computer (CORDIC)-based neuron architecture (RECON) which can be configured to compute both multiply-accumulate (MAC) and non-linear activation function (AF) operations. The CORDIC-based architecture uses linear and trigonometric relationships to realize MAC and AF operations respectively. The proposed design is synthesized and verified at 45nm technology using Cadence Virtuoso for all physical parameters. Implementation of the signed fixed-point 8-bit MAC using our design, shows 60% less area, latency, and power product (ALP) and shows improvement by 38% in area, 27% in power dissipation, and 15% in latency with respect to the state-of-the-art MAC design. Further, Monte-Carlo simulations for process-variations and device-mismatch are performed for both the proposed model and the state-of-the-art to evaluate expectations of functions of randomness in dynamic power variation. The dynamic power variation for our design shows that worst-case mean is 189.73μW which is 63% of the state-of-the-art. INDEX TERMS AF, CORDIC, configurable architecture, MAC, neural network.
In the last decade, the public health problem is the primary consciousness worldwide, and cancer is especially the central issue. Further, skin cancer comes in the top-3 of the world's most common cancer. We have proposed an efficient convolutional neural network (CNN) model that identifies skin cancer problems accurately. Although dataset HAM10K is used for the classification problem, its samples for each class are highly imbalanced and therefore are accountable for lower training accuracy. The AlexNet model is customized for the HAM10K data classification to address this problem. In addition, this work has presented an activation function that overcomes the vanishing gradient problem, and it is verified using the used dataset at multiple benchmark architectures. The results show better accuracy compared to the state-of-the-art activation function. Our customized CNN architecture with the proposed activation function has 98.20% accuracy for HAM10K, which is better than any other state-of-the-art models currently present. Further, precision, recall, and F-score results are also improved, which are also 98.20%.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.