TY - JOUR
T1 - ApGAN
T2 - Approximate GAN for Robust Low Energy Learning from Imprecise Components
AU - Roohi, Arman
AU - Sheikhfaal, Shadi
AU - Angizi, Shaahin
AU - Fan, Deliang
AU - Demara, Ronald F.
N1 - Funding Information:
We would like to acknowledge and thank the Advanced Research Computing Center at the University of Central Florida for provision of computing resources used herein. This work was supported in part by the Center for Probabilistic Spin Logic for Low-Energy Boolean and Non-Boolean Computing (CAPSL), one of the Nanoelectronic Computing Research (nCORE) Centers as task 2759.006, a Semiconductor Research Corporation (SRC) program sponsored by the NSF through CCF-1739635.
Publisher Copyright:
© 1968-2012 IEEE.
PY - 2020/3/1
Y1 - 2020/3/1
N2 - A Generative Adversarial Network (GAN) is an adversarial learning approach which empowers conventional deep learning methods by alleviating the demands of massive labeled datasets. However, GAN training can be computationally-intensive limiting its feasibility in resource-limited edge devices. In this paper, we propose an approximate GAN (ApGAN) for accelerating GANs from both algorithm and hardware implementation perspectives. First, inspired by the binary pattern feature extraction method along with binarized representation entropy, the existing Deep Convolutional GAN (DCGAN) algorithm is modified by binarizing the weights for a specific portion of layers within both the generator and discriminator models. Further reduction in storage and computation resources is achieved by leveraging a novel hardware-configurable in-memory addition scheme, which can operate in the accurate and approximate modes. Finally, a memristor-based processing-in-memory accelerator for ApGAN is developed. The performance of the ApGAN accelerator on different data-sets such as Fashion-MNIST, CIFAR-10, STL-10, and celeb-A is evaluated and compared with recent GAN accelerator designs. With almost the same Inception Score (IS) to the baseline GAN, the ApGAN accelerator can increase the energy-efficiency by sim 28.6×∼28.6× achieving 35-fold speedup compared with a baseline GPU platform. Additionally, it shows 2.5× and 5.8× higher energy-efficiency and speedup over CMOS-ASIC accelerator subject to an 11 percent reduction in IS.
AB - A Generative Adversarial Network (GAN) is an adversarial learning approach which empowers conventional deep learning methods by alleviating the demands of massive labeled datasets. However, GAN training can be computationally-intensive limiting its feasibility in resource-limited edge devices. In this paper, we propose an approximate GAN (ApGAN) for accelerating GANs from both algorithm and hardware implementation perspectives. First, inspired by the binary pattern feature extraction method along with binarized representation entropy, the existing Deep Convolutional GAN (DCGAN) algorithm is modified by binarizing the weights for a specific portion of layers within both the generator and discriminator models. Further reduction in storage and computation resources is achieved by leveraging a novel hardware-configurable in-memory addition scheme, which can operate in the accurate and approximate modes. Finally, a memristor-based processing-in-memory accelerator for ApGAN is developed. The performance of the ApGAN accelerator on different data-sets such as Fashion-MNIST, CIFAR-10, STL-10, and celeb-A is evaluated and compared with recent GAN accelerator designs. With almost the same Inception Score (IS) to the baseline GAN, the ApGAN accelerator can increase the energy-efficiency by sim 28.6×∼28.6× achieving 35-fold speedup compared with a baseline GPU platform. Additionally, it shows 2.5× and 5.8× higher energy-efficiency and speedup over CMOS-ASIC accelerator subject to an 11 percent reduction in IS.
KW - Generative adversarial network
KW - hardware mapping
KW - in-memory processing platform
KW - neural network acceleration
UR - http://www.scopus.com/inward/record.url?scp=85079629446&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85079629446&partnerID=8YFLogxK
U2 - 10.1109/TC.2019.2949042
DO - 10.1109/TC.2019.2949042
M3 - Article
AN - SCOPUS:85079629446
VL - 69
SP - 349
EP - 360
JO - IEEE Transactions on Computers
JF - IEEE Transactions on Computers
SN - 0018-9340
IS - 3
M1 - 8880521
ER -