Demonstration of Applications in Computer Vision and NLP on Ultra Power-Efficient CNN Domain Specific Accelerator with 9.3TOPS/Watt

Sun, Baoqing; Yang, Lin; Zhang, Wenhan; Dong, Patrick; Young, Charles; Dong, Jason; Lin, Michael

doi:10.1109/icmew.2019.00115

Cited by 5 publications

(5 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The run times of CNN-MIMO can be further accelerated by implementing the network in general-purpose hardware such as FPGA. For example, domain-specific architectures have been implemented in [46] for AlexNet [43] and VGG-16 for real-time image classification with 194 GOP/s (billions of fixed-point OPerations per second) and consuming only 300 mW. These promising results encourage us to develop more energy-efficient DL approaches for the problems in communications systems.…”

Section: Numerical Simulationsmentioning

confidence: 99%

Hybrid Precoding for Multiuser Millimeter Wave Massive MIMO Systems: A Deep Learning Approach

Elbir

Papazafeiropoulos

2020

IEEE Trans. Veh. Technol.

135

View full text Add to dashboard Cite

In multi-user millimeter wave (mmWave) multipleinput-multiple-output (MIMO) systems, hybrid precoding is a crucial task to lower the complexity and cost while achieving a sufficient sum-rate. Previous works on hybrid precoding were usually based on optimization or greedy approaches. These methods either provide higher complexity or have sub-optimum performance. Moreover, the performance of these methods mostly relies on the quality of the channel data. In this work, we propose a deep learning (DL) framework to improve the performance and provide less computation time as compared to conventional techniques. In fact, we design a convolutional neural network for MIMO (CNN-MIMO) that accepts as input an imperfect channel matrix and gives the analog precoder and combiners at the output. The procedure includes two main stages. First, we develop an exhaustive search algorithm to select the analog precoder and combiners from a predefined codebook maximizing the achievable sum-rate. Then, the selected precoder and combiners are used as output labels in the training stage of CNN-MIMO where the input-output pairs are obtained. We evaluate the performance of the proposed method through numerous and extensive simulations and show that the proposed DL framework outperforms conventional techniques. Overall, CNN-MIMO provides a robust hybrid precoding scheme in the presence of imperfections regarding the channel matrix. On top of this, the proposed approach exhibits less computation time with comparison to the optimization and codebook based approaches.

show abstract

Section: Numerical Simulationsmentioning

confidence: 99%

Hybrid Precoding for Multiuser Millimeter Wave Massive MIMO Systems: A Deep Learning Approach

Elbir

Papazafeiropoulos

2020

IEEE Trans. Veh. Technol.

135

View full text Add to dashboard Cite

show abstract

“…In this study, a well-known deep learning algorithm CNN is used. CNN algorithms are applied in many different areas such as natural language processing (NLP) [13,14], biomedical processing [15,16], especially image [17,18] and audio processing [19,20]. It is a deep learning algorithm that has the best classification success, especially in the field of image processing.…”

Section: Classificationmentioning

confidence: 99%

Deep Learning Method for Handwriting Recognition

Erdoğan

Tümer

2021

MANAS Journal of Engineering

View full text Add to dashboard Cite

The advancement of technology nowadays resulted into documents, such as forms and petitions, being filled out in computer and digital environment. Yet in some cases, documents are still preserved in traditional style, on print. Due to its distinct proportions, however, its storage, sharing and filing has become a complication. The relocation of these written documents to digital environment is therefore of great significance. In this view, this study aims to explore methodologies of digitizing handwritten documents. In this study, the documents converted to image format were pre-processed using image processing methods. These operations include dividing lines of the document into image format, dividing into words which then divided into characters, and finally, a classification operation on the characters. As classification phase, one of the deep learning methods is the Convolution Neural Network method is used in image recognition. The model was trained using the EMNIST dataset, and in the character, dataset created from the documents at hand. The dataset created had a success rate of 87.81%. Characters classified as finishers are sequentially combined and the document is transferred to the computer afterwards.

show abstract

“…GnetFC-v1 [13,16] is an image classification model specially designed for the CNN accelerator chip. It can be also used for NLP and multi-modal tasks [22,24,23]. The motivation of GnetFC-v1 model is to reduce the CPU load of image classification models by replacing the FC (Fully Connected) layers with the 3x3 convolutional layers.…”

Section: From Gnetfc-v1 To Gnetfc-v2mentioning

confidence: 99%

“…The most recent chip has the peak power of only 224mW [14]. These low-power CNN accelerators are used in computer vision tasks [16,24,13], and also in NLP (Natural Language Processing) tasks [18,17,20,21,22,11], and extended into tabular data machine learning [25] and multimodal tasks [23].…”

Section: Introductionmentioning

confidence: 99%

GnetDet: Object Detection Optimized on a 224mW CNN Accelerator Chip at the Speed of 106FPS

Sun¹,

Zhang²,

Su³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Object detection is widely used on embedded devices. With the wide availability of CNN (Convolutional Neural Networks) accelerator chips, the object detection applications are expected to run with low power consumption, and high inference speed. In addition, the CPU load is expected to be as low as possible for a CNN accelerator chip working as a co-processor with a host CPU. In this paper, we optimize the object detection model on the CNN accelerator chip by minimizing the CPU load. The resulting model is called GnetDet. The experimental result shows that the GnetDet model running on a 224mW chip achieves the speed of 106FPS with excellent accuracy.

show abstract

Demonstration of Applications in Computer Vision and NLP on Ultra Power-Efficient CNN Domain Specific Accelerator with 9.3TOPS/Watt

Cited by 5 publications

References 0 publications

Hybrid Precoding for Multiuser Millimeter Wave Massive MIMO Systems: A Deep Learning Approach

Hybrid Precoding for Multiuser Millimeter Wave Massive MIMO Systems: A Deep Learning Approach

Deep Learning Method for Handwriting Recognition

GnetDet: Object Detection Optimized on a 224mW CNN Accelerator Chip at the Speed of 106FPS

Contact Info

Product

Resources

About