Deep Learning in the Era of Edge Computing: Challenges and Opportunities

Zhang, Mi; Zhang, Faen; Lane, Nicholas D.; Shu, Yuanchao; Zeng, Xiao; Fang, Biyi; Shen, Yan; Xu, Hui

doi:10.1002/9781119551713.ch3

Cited by 31 publications

(16 citation statements)

References 24 publications

(20 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Deep learning is computationally expensive compared to traditional machine learning, needing a huge amount of memory and processing resources, and it is difficult to adapt to new situations. It is difficult to put into words and is not totally understood [ 55 , 56 ]. As a result, we only talked about the applicability of the self paced learning technique to scRNA-seq data in the traditional machine learning model in this study.…”

Section: Discussionmentioning

confidence: 99%

Single Cell Self-Paced Clustering with Transcriptome Sequencing Data

Zhao

Chen

et al. 2022

IJMS

View full text Add to dashboard Cite

Single cell RNA sequencing (scRNA-seq) allows researchers to explore tissue heterogeneity, distinguish unusual cell identities, and find novel cellular subtypes by providing transcriptome profiling for individual cells. Clustering analysis is usually used to predict cell class assignments and infer cell identities. However, the performance of existing single-cell clustering methods is extremely sensitive to the presence of noise data and outliers. Existing clustering algorithms can easily fall into local optimal solutions. There is still no consensus on the best performing method. To address this issue, we introduce a single cell self-paced clustering (scSPaC) method with F-norm based nonnegative matrix factorization (NMF) for scRNA-seq data and a sparse single cell self-paced clustering (sscSPaC) method with l21-norm based nonnegative matrix factorization for scRNA-seq data. We gradually add single cells from simple to complex to our model until all cells are selected. In this way, the influences of noisy data and outliers can be significantly reduced. The proposed method achieved the best performance on both simulation data and real scRNA-seq data. A case study about human clara cells and ependymal cells scRNA-seq data clustering shows that scSPaC is more advantageous near the clustering dividing line.

show abstract

Section: Discussionmentioning

confidence: 99%

Single Cell Self-Paced Clustering with Transcriptome Sequencing Data

Zhao

Chen

et al. 2022

IJMS

View full text Add to dashboard Cite

show abstract

“…In the era of edge computing, equipping the edge devices with AI powered deep learning models has attracted the attention of many researchers and companies for providing the real-time solutions for deployment. However, despite the edge computing advantages of low latency, scalability and privacy, the deployment of the DL-based models on the edge devices is still a major challenge in terms of computation, memory, and power consumption [62]. Typically, once the model is trained with the popular DL-based frameworks such as TensorFlow [12], PyTorch [14], Darknet [63] or Caffe [13].…”

Section: B Optimization For Edge Deploymentmentioning

confidence: 99%

Edge Deployment Framework of GuardBot for Optimized Face Mask Recognition With Real-Time Inference Using Deep Learning

Manzoor¹,

Kim

Joo

et al. 2022

IEEE Access

View full text Add to dashboard Cite

Deep learning based models on the edge devices have received considerable attention as a promising means to handle a variety of AI applications. However, deploying the deep learning models in the production environment with efficient inference on the edge devices is still a challenging task due to computation and memory constraints. This paper proposes a framework for the service robot named GuardBot powered by Jetson Xavier NX and presents a real-world case study of deploying the optimized face mask recognition application with real-time inference on the edge device. It assists the robot to detect whether people are wearing a mask to guard against COVID-19 and gives a polite voice reminder to wear the mask. Our framework contains dual-stage architecture based on convolutional neural networks with three main modules that employ (1) MTCNN for face detection, (2) our proposed CNN model and seven transfer learning based custom models which are Inception-v3, VGG16, denseNet121, resNet50, NASNetMobile, XceptionNet, MobileNet-v2 for face mask classification, (3) TensorRT for optimization of all the models to speedup inference on the Jetson Xavier NX. Our study carries out several analysis based on the models' performance in terms of their frames per second, execution time and images per second. It also evaluates the accuracy, precision, recall & F1-score and makes the comparison of all models before and after optimization with a main focus on high throughput and low latency. Finally, the framework is deployed on a mobile robot to perform experiments in both outdoor and multi-floor indoor environments with patrolling and non-patrolling modes. Compared to other state-of-the-art models, our proposed CNN model for face mask recognition based on the classification obtains 94.5%, 95.9% and 94.28% accuracy on training, validation and testing datasets respectively which is better than MobileNet-v2, Xception and InceptionNet-v3 while it achieves highest throughput and lowest latency than all other models after optimization at different precision levels.

show abstract

“…A relevant challenge, worthy of further consideration, is to understand the performance trade-offs at scale of combining a variety of learning paradigms such as Reinforcement Learning [155], Deep Learning [59], Online Learning [44], Stream Learning [84], Lifelong Learning [52], Transfer Learning [44], Federated Learning [9], Distributed Learning [164,166], Multi-task Learning [168], and others.…”

Section: How To Combine Machine Learning Paradigms To Leverage the Ma...mentioning

confidence: 99%