Latency in Cloud-Based Interactive Streaming Content

Sharp, Ron

doi:10.1002/bltj.21545

Cited by 13 publications

(3 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The target Service-Level Objective (SLO) for round-trip service latency depends on the nature of the application or website using RaaS services. As a rule of thumb, we consider in this evaluation that a median latency below 300 ms (not accounting the latency to and from the data center hosting the RaaS services) and never exceeding twice that value should comply with typical SLOs for online services [69] 8 . We use the MovieLens dataset ml-20m [44,47] as our experimental workload.…”

Section: Discussionmentioning

confidence: 99%

PProx

Rosinosky¹,

Silva

Mokhtar

et al. 2021

Proceedings of the 22nd International Middleware Conference

View full text Add to dashboard Cite

We present PProx, a system preventing recommendation-as-a-service (RaaS) providers from accessing sensitive data about the users of applications leveraging their services. PProx does not impact recommendations accuracy, is compatible with arbitrary recommendation algorithms, and has minimal deployment requirements. Its design combines two proxying layers directly running inside SGX enclaves at the RaaS provider side. These layers transparently pseudonymize users and items and hide links between the two, and PProx privacy guarantees are robust even to the corruption of one of these enclaves. We integrated PProx with the Harness recommendation engine and evaluated it on a 27-node cluster. Our results indicate its ability to withstand a high number of requests with low end-to-end latency, horizontally scaling up to match increasing workloads of recommendations.CCS Concepts • Security and privacy;

show abstract

Section: Discussionmentioning

confidence: 99%

PProx

Rosinosky¹,

Silva

Mokhtar

et al. 2021

Proceedings of the 22nd International Middleware Conference

View full text Add to dashboard Cite

show abstract

“…A common approach to overcome the computational resources is to leverage cloud computing [50] in which data of IoT sensors is transmitted from the source to a centralized location. It introduces challenges related to the latency [51], scalability and privacy [52] that are critical for many applications. For example, camera frames of AV require to be swiftly processed for detection and obstacle avoidance or voice commands given to a robot need to be instantly parsed for returning the quick response to the user's query.…”

Section: A Edge Vs Cloud Computingmentioning

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

“…), RTT values below 150 ms are shown to be acceptable, while for more interactive applications, such as gaming, 80 ms RTT is found as an upper limit for usability. In [5] it is shown that roughly half of the total round trip time is caused by packetization and propagating data over the network.…”

Section: Introductionmentioning

confidence: 99%

Network latency hiding in thin client systems through server-centric speculative display updating

Vankeirsbilck¹,

Simoens²,

Turck³

et al. 2014

Journal of Network and Computer Applications

View full text Add to dashboard Cite

The widespread availability of cloud computing services has revitalized interest in the thin client computing paradigm, in which application logic is executing on a remote server, typically hosted in a cloud computing infrastructure. The user interacts with a local viewer, that forwards the user events over the network to the server and accepts the returned graphical updates. An important challenge for this approach consists of the fact that at least one network round-trip time is required to present the application output that results from the user's actions. In this paper a novel speculative display update mechanism is proposed to hide the network latency from the user by speculatively updating the screen without awaiting the server response. The mechanism relies on online server side profiling of the graphical output caused by user events, based on which a finite-state model is constructed capturing the graphical behaviour of the application. Experiments with a text editor show that, once the application model is learned, speculative responses are displayed within 40 ms for over 80% of the user events, with an accuracy exceeding 70%.

show abstract

Latency in Cloud-Based Interactive Streaming Content

Cited by 13 publications

References 2 publications

PProx

PProx

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

Network latency hiding in thin client systems through server-centric speculative display updating

Contact Info

Product

Resources

About