Internet of video things in 2030: A world with many cameras

Mohan, Ashwin; Gauen, Kent; Lu, Yung-Hsiang; Li, Wei Wayne; Chen, Xuemin

doi:10.1109/iscas.2017.8050296

Cited by 48 publications

(20 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…However, making groups of visually similar categories is not a well-defined operation that can be obtained using a fixed threshold. For example, if we set the threshold to the value 1 10 (softmax value obtained from untrained DNN, assuming M = 10), then a category with an ASL of 0.099 will not be selected. However, a category with an ASL of 0.101 will always be selected.…”

Section: Similarity Metric: Averaged Softmax Likelihoodmentioning

confidence: 99%

Modular Neural Networks for Low-Power Image Classification on Embedded Devices

Goel

Aghajanzadeh

Tung

et al. 2020

ACM Trans. Des. Autom. Electron. Syst.

Self Cite

View full text Add to dashboard Cite

Embedded devices are generally small, battery-powered computers with limited hardware resources. It is difficult to run deep neural networks (DNNs) on these devices, because DNNs perform millions of operations and consume significant amounts of energy. Prior research has shown that a considerable number of a DNN's memory accesses and computation are redundant when performing tasks like image classification. To reduce this redundancy and thereby reduce the energy consumption of DNNs, we introduce the Modular Neural Network Tree architecture. Instead of using one large DNN for the classifier, this architecture uses multiple smaller DNNs (called modules) to progressively classify images into groups of categories based on a novel visual similarity metric. Once a group of categories is selected by a module, another module then continues to distinguish among the similar categories within the selected group. This process is repeated over multiple modules until we are left with a single category. The computation needed to distinguish dissimilar groups is avoided, thus reducing redundant operations, memory accesses, and energy. Experimental results using several image datasets reveal the effectiveness of our proposed solution to reduce memory requirements by 50% to 99%, inference time by 55% to 95%, energy consumption by 52% to 94%, and the number of operations by 15% to 99% when compared with existing DNN architectures, running on two different embedded systems: Raspberry Pi 3 and Raspberry Pi Zero. CCS Concepts: • Computing methodologies → Neural networks; Computer vision; • Computer systems organization → Embedded systems; • Hardware → Power and energy;

show abstract

Section: Similarity Metric: Averaged Softmax Likelihoodmentioning

confidence: 99%

Modular Neural Networks for Low-Power Image Classification on Embedded Devices

Goel

Aghajanzadeh

Tung

et al. 2020

ACM Trans. Des. Autom. Electron. Syst.

Self Cite

View full text Add to dashboard Cite

show abstract

“…IDC estimated that global IoT spending will reach $1.29 trillion by 2020 [57], [58]. In [59] authors predicts that the number of connected cameras to the internet will reach 100 billion in 20130 (Internet of video things). Fig.…”

Section: A History Of the Internet Of Thingsmentioning

confidence: 99%

Systematic Review of Existing IoT Architectures Security and Privacy Issues and Concerns

Alshohoumi¹,

Sarrab²,

AlHamadani³

et al. 2019

IJACSA

View full text Add to dashboard Cite

Internet of things (IoT) has become one of the most prominent technologies that the world has been witnessing nowadays. It provides great solutions to humanity in many significant fields of life. IoT refers to a collection of sensors or object in the universe with the capability of communicating with each other through the internet without human intervention. Currently, there is no standard IoT architecture. As it is in its infancy, IoT is surrounded by numerous security and privacy concerns. Thus, to avoid such concerns that may hinder its deployment, an IoT architecture has to be carefully designed to incorporate security and privacy solutions. In this paper, a systematic literature review was conducted to trace the evolvement of IoT architectures from its initial development in 2008 until 2018. The Comparison among these architectures is based on terms of the architectural stack, covered issues, the technology used and considerations of security and privacy aspects. The findings of the review show that the initial IoT architectures did not provide a comprehensive meaning for IoT that describe its nature, whereas the recent IoT architectures convey a comprehensive meaning of IoT, starting from data collection, followed by data transmission and processing, and ending with data dissemination. Moreover, the findings reveal that IoT architecture has evolved gradually across the years, through improving architecture stack with new solutions to mitigate IoT challenges such as scalability, interoperability, extensibility, management, etc. with lack consideration of security solutions. The findings disclose that none of the discussed IoT architectures considers privacy concerns, which indeed considered as a critical factor of IoT sustainability and success. Therefore, there is an inevitable need to consider security and privacy solutions when designing IoT architecture.

show abstract

“…In contrast, network cameras can continuously acquire visual data at fixed locations, and the data can be used for comparison [6]. Taking an example from Houston traffic cameras during the 2017 Hurricane Harvey, Figure 2 Prior work has explored using network cameras for public safety [7].…”

Section: A Images and Videos For Emergency Response And Public Safetymentioning

confidence: 99%

Cross-referencing social media and public surveillance camera data for disaster response

Surakitbanharn

Yau

Wang

et al. 2018

2018 IEEE International Symposium on Technologies for Homeland Security (HST)

Self Cite

View full text Add to dashboard Cite

Physical media (like surveillance cameras) and social media (like Instagram and Twitter) may both be useful in attaining on-the-ground information during an emergency or disaster situation. However, the intersection and reliability of both surveillance cameras and social media during a natural disaster are not fully understood. To address this gap, we tested whether social media is of utility when physical surveillance cameras went off-line during Hurricane Irma in 2017. Specifically, we collected and compared geo-tagged Instagram and Twitter posts in the state of Florida during times and in areas where public surveillance cameras went off-line. We report social media content and frequency and content to determine the utility for emergency managers or first responders during a natural disaster.

show abstract

Internet of video things in 2030: A world with many cameras

Cited by 48 publications

References 6 publications

Modular Neural Networks for Low-Power Image Classification on Embedded Devices

Modular Neural Networks for Low-Power Image Classification on Embedded Devices

Systematic Review of Existing IoT Architectures Security and Privacy Issues and Concerns

Cross-referencing social media and public surveillance camera data for disaster response

Contact Info

Product

Resources

About