Filters








31,387 Hits in 10.9 sec

CNN based high performance computing for real time image processing on GPU

Sasanka Potluri, Alireza Fasih, Laxminand Kishore Vutukuru, Fadi Al Machot, Kyandoghere Kyamakya
2011 Proceedings of the Joint INDS'11 & ISTET'11  
The inherent massive parallelism of CNN along with GPUs makes it an advantage for high performance computing platform [2].  ...  In real time applications the processing time is considered as a big obstacle for its implementations. A High Performance Computing (HPC) platform is necessary in order to solve this problem.  ...  Contrast enchamcement operation For real time image processing applications the performance of CNN is evaluated for the contrast enhancement.  ... 
doi:10.1109/inds.2011.6024781 fatcat:khbsess6szdf3n3c5eyp3jlcfm

CNN Based High Performance Computing for Real Time Image Processing on GPU [chapter]

Sasanka Potluri, Alireza Fasih, Laxminand Kishore Vutukuru, Fadi Al Machot, Kyandoghere Kyamakya
2012 Studies in Computational Intelligence  
The inherent massive parallelism of CNN along with GPUs makes it an advantage for high performance computing platform [2].  ...  In real time applications the processing time is considered as a big obstacle for its implementations. A High Performance Computing (HPC) platform is necessary in order to solve this problem.  ...  Contrast enchamcement operation For real time image processing applications the performance of CNN is evaluated for the contrast enhancement.  ... 
doi:10.1007/978-3-642-24806-1_20 fatcat:nootjvvgivhnfeec5zpirpy6ra

Deep Learning-Based Multiple Object Visual Tracking on Embedded System for IoT and Mobile Edge Computing Applications [article]

Beatriz Blanco-Filgueira, Daniel García-Lesta, Mauro Fernández-Sanjurjo, Víctor M. Brea, Paula López
2018 arXiv   pre-print
This proposal performs low-power and real time deep learning-based multiple object visual tracking implemented on an NVIDIA Jetson TX2 development kit.  ...  for IoT and mobile edge computing applications due to their high power consumption.  ...  The results of the last years depict a hopeful prospect for image processing using CNNs.  ... 
arXiv:1808.01356v1 fatcat:3inib27h25bsvl6n4rulgqhtpq

GPU-Based Embedded Intelligence Architectures and Applications

Li Minn Ang, Kah Phooi Seng
2021 Electronics  
This paper present contributions to the state-of-the art for graphics processing unit (GPU-based) embedded intelligence (EI) research for architectures and applications.  ...  This paper gives a comprehensive review and representative studies of the emerging and current paradigms for GPU-based EI with the focus on the architecture, technologies and applications: (1) First, the  ...  Their proposed GPU-based embedded image recognition approach used the NVIDIA GPU to accelerate the computations for an immune CNN to perform the image recognition.  ... 
doi:10.3390/electronics10080952 fatcat:paubm2sevbhixi2in63ayflmti

A comparison study between MLP and convolutional neural network models for character recognition

S. Ben Driss, M. Soua, R. Kachouri, M. Akil, Nasser Kehtarnavaz, Matthias F. Carlsohn
2017 Real-Time Image and Video Processing 2017  
Based on our experimentations, we demonstrate that the used real-time CNN is 2x more relevant than MLP when classifying characters.  ...  Moreover, the convolutional neural network (CNN), is gaining nowadays a lot of popularity for its high performance.  ...  Real time evaluation We evaluate the learning and classification time on GPU on the chars74k dataset. 20 Table 2 shows the total processing time for learning and classification on all characters of  ... 
doi:10.1117/12.2262589 fatcat:rsesq6kzwvcebphpuh7t5duabq

GBCNN: A Full GPU Based Batch Multi-task Cascaded Convolutional Networks

Shijie Li, Yong Dou, Jinwei Xu, Ke Yang, Rongchun Li
2019 IEEE Access  
However, it has difficulty in predicting faces among the big scale images in real time due to its three stages cascade architecture with less optimization.  ...  In this paper, we propose a full GPU-based batch multi-task cascade convolutional network which is carefully designed and optimized in each step to gain a superior speed performance.  ...  At that time, the detector achieves a nearly real-time performance.  ... 
doi:10.1109/access.2019.2894589 fatcat:a7bzuxt6trelfgfy4xuky5acmq

DeepPicar: A Low-cost Deep Neural Network-based Autonomous Car [article]

Michael G. Bechtel, Elise McEllhiney, Minje Kim, Heechul Yun
2018 arXiv   pre-print
We also systematically compare other contemporary embedded computing platforms using the DeepPicar's CNN-based real-time control workload.  ...  We find that all tested platforms, including the Pi 3, are capable of supporting the CNN-based real-time control, from 20 Hz up to 100 Hz, depending on hardware platform.  ...  The Titan Xp and Jetson TX2 used for this research were donated by the NVIDIA Corporation.  ... 
arXiv:1712.08644v4 fatcat:3cdrhgxvlrf4fniuss7jt7qtny

A comparison of CNN-based face and head detectors for real-time video surveillance applications

Le Thanh Nguyen-Meidine, Eric Granger, Madhu Kiran, Louis-Antoine Blais-Morin
2017 2017 Seventh International Conference on Image Processing Theory, Tools and Applications (IPTA)  
Results suggest that, although CNN architectures can achieve a very high level of accuracy compared to traditional detectors, their computational cost can represent a limitation for many practical real-time  ...  Single pass and region-based architectures are reviewed and compared empirically to baseline techniques according to accuracy and to time and memory complexity on images from several challenging datasets  ...  The current processing time of SSD is 12.5 fps on 720p images, and 10.3 fps on 1080p images. V. CONCLUSION Computational complexity is an important problem for practical CNN application.  ... 
doi:10.1109/ipta.2017.8310113 dblp:conf/ipta/Nguyen-MeidineG17 fatcat:2qp3wkxfb5htbo2yh6r3u255sm

Optimizing CNN-based Hyperspectral Image Classification on FPGAs [article]

Shuanglong Liu, Ringo S.W. Chu, Xiwei Wang, Wayne Luk
2019 arXiv   pre-print
Hyperspectral image (HSI) classification has been widely adopted in applications involving remote sensing imagery analysis which require high classification accuracy and real-time processing speed.  ...  This paper proposes a novel CNN-based algorithm for HSI classification which takes into account hardware efficiency.  ...  These former accelerators achieve real time processing speed but they do not achieve high classification accuracy and therefore are not favoured over CNN-based methods.  ... 
arXiv:1906.11834v1 fatcat:arcbhexooja6hhmm4j5z4sgbei

DroNet: Efficient convolutional neural network detector for real-time UAV applications

Christos Kyrkou, George Plastiras, Theocharis Theocharides, Stylianos I. Venieris, Christos-Savvas Bouganis
2018 2018 Design, Automation & Test in Europe Conference & Exhibition (DATE)  
embedded processing platform suitable for deployment on UAVs.  ...  This paper therefore, explores the trade-offs involved in the development of a single-shot object detector based on deep convolutional neural networks (CNNs) that can enable UAVs to perform vehicle detection  ...  ACKNOWLEDGMENT Christos Kyrkou gratefully acknowledges the support of NVIDIA Corporation with the donation of the Titan Xp GPU used for this research.  ... 
doi:10.23919/date.2018.8342149 dblp:conf/date/KyrkouPTVB18 fatcat:73nr4wvsnvcehjzgphiil3n5ey

Enhancement of Plant Disease Detection Framework using Cloud Computing and GPU Computing

2019 International Journal of Engineering and Advanced Technology  
GPUs are very useful in high performance computing. With the emerging new trend of cloud environments with GPU instances are now gaining popularity in many real time applications.  ...  This research paper focuses on the applications of GPU Computing within cloud and to study the performance of GPU image processing and Normal CPU image processing with respect to plant disease detection  ...  APPLICATIONS OF GPU WITH CLOUD COMPUTING Save Money with GPU Cloud Computing Normal CPU takes a long time when working with image processing on high quality images .The processing time is so high that  ... 
doi:10.35940/ijeat.a9541.109119 fatcat:umdhedd6d5c4zmdkthkbmyygse

Fast CNN Stereo Depth Estimation through Embedded GPU Devices

Cristhian A. Aguilera, Cristhian Aguilera, Cristóbal A. Navarro, Angel D. Sappa
2020 Sensors  
Current CNN-based stereo depth estimation models can barely run under real-time constraints on embedded graphic processing unit (GPU) devices.  ...  In our experiments, we achieve real-time inference speed, in the range of 5–32 ms, for 1216 × 368 input stereo images on the Jetson TX2, Jetson Xavier, and Jetson Nano embedded devices.  ...  Therefore, highly accurate models designed for high-end desktop GPUs today cannot run under real-time constraints on current embedded GPU devices.  ... 
doi:10.3390/s20113249 pmid:32517319 fatcat:ybuq6blekfdyjba6olnzgynb2m

Compact Convolutional Neural Network Cascade for Face Detection [article]

Ilya Kalinovskii, Vladimir Spitsyn
2015 arXiv   pre-print
Because of high computational efficiency, our detector can processing 4K Ultra HD video stream in real time (up to 27 fps) on mobile platforms (Intel Ivy Bridge CPUs and Nvidia Kepler GPUs) in searching  ...  Many algorithms achieve a high quality face detection, but at the cost of high computational complexity. This restricts their application in the real-time systems.  ...  Due to the natural parallelism, a small number of cascade stages and lowlevel optimization, it is capable of processing real-time 4K Ultra HD video stream on mobile GPU when searching for faces with the  ... 
arXiv:1508.01292v3 fatcat:xqze7hjeuncqroybgcarj22bsi

Deep Learning-Based Real-Time Multiple-Object Detection and Tracking from Aerial Imagery via a Flying Robot with GPU-Based Embedded Devices

Hossain, Lee
2019 Sensors  
These are suitable for real-time onboard computing power on small flying drones with limited space.  ...  We also introduce an effective target tracking approach for moving objects. The algorithm for tracking moving objects is based on the extension of simple online and real-time tracking.  ...  I would also like to pay a deep sense of gratitude to all CAIAS (Center for Artificial Intelligence and Autonomous System) lab members for their support and CAIAS lab for providing me all the facilities  ... 
doi:10.3390/s19153371 fatcat:htf3ilkn3ndrxjzoaos6vj6fc4

The impact of the soft errors in convolutional neural network on GPUs: Alexnet as case study

Khalid Adam, Izzeldin I. Mohd, Younis M. Younis
2021 Procedia Computer Science  
Meanwhile, the vulnerability of CNN model to soft errors (e.g., caused by radiation induced) rapidly increases, thus reliability is crucial especially in real-time system.  ...  Meanwhile, the vulnerability of CNN model to soft errors (e.g., caused by radiation induced) rapidly increases, thus reliability is crucial especially in real-time system.  ...  CNNs have high massive parallelism structure and require high computational capabilities such as Graphics Processing Units (GPUs).  ... 
doi:10.1016/j.procs.2021.02.012 fatcat:jwloam4rcbhbpncybbdmwwszzi
« Previous Showing results 1 — 15 out of 31,387 results