Author:
Anjankar Shubham,Hemant Gillurkar ,Joshi Pankaj,Dwaramwar Pravin
Abstract
The demand for Deep Learning applications in resource constrained devices is booming in recent years. Theuse of Deep Neural Network (DNN) is the leading method in these applications which has error resilient nature.This allows the use of Approximate Computing for efficient computation to leverage efficiency accuracy tradeoff by replacing Approximate Multiplier in place of exact multipliers. In this paper we proposed ApproximateCompressors and compared them in different cases of 8 bit integer dadda multipliers in terms of the Error metricsand accuracy in real life object classification applications. The approximate multipliers are designed using differentcompressors and used to perform multiplication in ResNet. We have proposed two approximate compressorsdesigns Design 1 and Design 2.The proposed 4:2 compressors design shows the more correct outputs and lessWorst Case Relative Error (WCRE) in the range of 2-16. Our proposed 4:2 compressor Design1 is utilized in themodified Reduction circuitry of dadda multiplier and shows the accuracy of 81.6 % for DNN application
Publisher
Perpetual Innovation Media Pvt. Ltd.
Reference17 articles.
1. Aktas, K., Ignjatovic, V., Ilic, D., Marjanovic, M., and Anbarjafari, G. 2022. Deep convolutional neural networks for detection of abnormalities in chest X-rays trained on the very large dataset. Signal, Image and Video Processing.
2. Ansari, M. S., Mrazek, V., Cockburn, B. F., Sekanina, L., Vasicek, Z., and Han, J. 2020. Improving the Accuracy and Hardware Efficiency of Neural Networks Using Approximate Multipliers. IEEE Transactions on Very Large Scale Integration (VLSI) Systems 28, 2(2), 317–328.
3. Asakawa, E., Kaneko, N., Hasegawa, D., and Shirakawa, S. 2022. Evaluation of text-to-gesture generation model using convolutional neural network. Neural Networks 151,365–375.
4. Brodley, C. E., Rebbapragada, U., Small, K., and Wallace, B. 2012. Challenges and Opportunities in Applied Machine Learning. AI Magazine 33, 1 (mar 15), 11–24.
5. ehw fit. 2020. Github - ehw-fit/tf-approximate: Approximate layers - TensorFlow extension. https://github.com/ehw-fit/tf-approximate. [Online; accessed 2022-10-19].