Mean field variational inference using bregman ADMM for distributed camera network

Babagholami-Mohamadabadi, Behnam; Yoon, Seonghye; Pavlović, Vladimir

doi:10.1145/2789116.2802656

Cited by 2 publications

(2 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Due to limited on-sensor resources, on-sensor computation, with attendant energy overhead, can be avoided for applications in which the sampling rate of biomedical signals is low, e.g., monitoring the patient's body temperature [11]. However, in some applications, on-sensor computation is beneficial and preferred over off-sensor computation due to one of the following reasons [248,249]. First, on-sensor computation may significantly reduce the transmission energy (and as a result the total energy consumption of the device) even though it imposes extra energy consumption for computation.…”

Section: On-sensor Computationmentioning

confidence: 99%

Addressing Security and Privacy Challenges in Internet of Things

Mosenia

2018

Preprint

View full text Add to dashboard Cite

Internet of Things (IoT), also referred to as the Internet of Objects, is envisioned as a holistic and transformative approach for providing numerous services. The rapid development of various communication protocols and miniaturization of transceivers along with recent advances in sensing technologies offer the opportunity to transform isolated devices into communicating smart things. Smart things, that can sense, store, and even process electrical, thermal, optical, chemical, and other signals to extract user-/environment-related information, have enabled services only limited by human imagination.Despite picturesque promises of IoT-enabled systems, the integration of smart things into the standard Internet introduces several security challenges because the majority of Internet technologies, communication protocols, and sensors were not designed to support IoT. Several recent research studies have demonstrated that launching security/privacy attacks against IoT-enabled systems, in particular wearable medical sensor (WMS)-based systems, may lead to catastrophic situations and life-threatening conditions. Therefore, security threats and privacy concerns in the IoT domain need to be proactively studied and aggressively addressed. In this thesis, we tackle several domain-specific security/privacy challenges associated with IoTenabled systems.We first target health monitoring systems that are one of the most widely-used types of IoT-enabled systems. We discuss and evaluate several energy-efficient schemes and algorithms, which significantly reduce total energy consumption of different implantable and wearable medical devices (IWMDs). The proposed schemes make continuous long-term health monitoring feasible while providing spare energy needed for data encryption. Furthermore, we present two energy-efficient protocols for implantable medical devices (IMDs), which are essential for data encryption: (i) a secure wakeup protocol I have had the honor and privilege to work under the supervision of Prof. Niraj K. Jha. who helped me take the first steps as a researcher. His guidance and support have helped me with academic research and writing from the first day of the Ph.D. program and have been very beneficial in shaping my Ph.D work. Apart from his immense wisdom and insight, his enthusiasm to explore new research directions and novel ideas has been a constant source of inspiration to me. His patience, kindness, understanding, and empathy towards his students are truly admirable. He has been extremely caring during my days of struggle. Although I have immensely benefited from his technical expertise, I still have so much to learn from his superior discipline and kind disposition. I also received much advice and guidance from Profs. Anand Raghunathan, Susmita Sur-Kolay, and Mehran Mozaffari-Kermani. Discussion with them were always an exceptional source of new ideas. Without their continuous support, encouragement, and inspiration, preparing this thesis would not have been feasible. I am deeply grateful for the guidance of...

show abstract

Section: On-sensor Computationmentioning

confidence: 99%

Addressing Security and Privacy Challenges in Internet of Things

Mosenia

2018

Preprint

View full text Add to dashboard Cite

show abstract

“…Early studies in distributed visual recognition systems acquired images and compressed locally at the camera nodes, and then transmitted to the base station which performs the specific analysis tasks. In some recent approaches [1,2,3,4], the visual features (e.g. SIFT) have been extracted locally in each camera and then compressed and transmitted to the base station for further analysis (e.g.…”

Section: Introductionmentioning

confidence: 99%

Multi-view task-driven recognition in visual sensor networks

Taalimi

Rahimpour

Liu

et al. 2017

2017 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

Nowadays, distributed smart cameras are deployed for a wide set of tasks in several application scenarios, ranging from object recognition, image retrieval, and forensic applications. Due to limited bandwidth in distributed systems, efficient coding of local visual features has in fact been an active topic of research. In this paper, we propose a novel approach to obtain a compact representation of high-dimensional visual data using sensor fusion techniques. We convert the problem of visual analysis in resource-limited scenarios to a multiview representation learning, and we show that the key to finding properly compressed representation is to exploit the position of cameras with respect to each other as a normbased regularization in the particular signal representation of sparse coding. Learning the representation of each camera is viewed as an individual task and a multi-task learning with joint sparsity for all nodes is employed. The proposed representation learning scheme is referred to as the multi-view task-driven learning for visual sensor network (MT-VSN). We demonstrate that MT-VSN outperforms state-of-the-art in various surveillance recognition tasks.

show abstract

Mean field variational inference using bregman ADMM for distributed camera network

Cited by 2 publications

References 2 publications

Addressing Security and Privacy Challenges in Internet of Things

Addressing Security and Privacy Challenges in Internet of Things

Multi-view task-driven recognition in visual sensor networks

Contact Info

Product

Resources

About