Artificial Vision on Edge IoT Devices: A Practical Case for 3D Data Classification

Wisultschew, Cristian; Otero, Andrés; Portilla, Jorge; Torre, Eduardo de la

doi:10.1109/dcis201949030.2019.8959857

Cited by 7 publications

(3 citation statements)

References 14 publications

(12 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Early applications of AV focused on object detection and counting, nowadays, AV is used to characterize a wide range of object properties, such as size, shape, size distribution, and aggregation states (Wisultschew et al, 2019). With AV, image analysis can be automated, which reduces the time and effort required for characterization; it can be adapted to different sample types and experimental conditions.…”

Section: Ecological Engineering and Environmental Technologymentioning

confidence: 99%

Characterization of Metallic Gold Nanoparticles in a Colloidal State by Artificial Vision

Carbajal Morán,

Galván Maldonado,

Márquez Camarena

2024

Ecol. Eng. Environ. Technol.

View full text Add to dashboard Cite

Gold nanoparticles in their colloidal state have different colors, and the equipment for their characterization, such as UV-Vis spectrophotometers, scanning electron microscopy (SEM), and transmission electron microscopy (TEM), has high costs. The research aimed to characterize metallic gold nanoparticles by artificial vision based on the color of the samples in the colloidal state. The sensor used for the sampling was a 50 MP triple-lens camera with the optical image stabilization (OIS) of a smartphone. The Vision Acquisition and Vision Assistant blocks in the NI LabVIEW platform were used to implement an artificial vision device. The camera interface was used to identify the color of each of the 10 samples of colloidal gold nanoparticles produced by the YAG laser and chemical reduction in 15 ml of deionized water. The characterization consisted of the determination of the size and concentration of the gold nanoparticles based on their color, which ranged from pink to red wine. As a result, the artificial vision device adequately identified the color of the metallic gold nanoparticles in a colloidal state with a certainty of more than 95%, allowing the nanoparticles to be adequately characterized. Therefore, it is concluded that artificial vision adequately characterized gold nanoparticles' wavelength, absorbance, diameter, and concentration.

show abstract

Section: Ecological Engineering and Environmental Technologymentioning

confidence: 99%

Characterization of Metallic Gold Nanoparticles in a Colloidal State by Artificial Vision

Carbajal Morán,

Galván Maldonado,

Márquez Camarena

2024

Ecol. Eng. Environ. Technol.

View full text Add to dashboard Cite

show abstract

“…A number of ML hardware accelerator platforms, such as Google's Edge TPU, have recently been developed, and a number of papers have evaluated them. For example, Wisultschew et al [5] studied and compared the performance and efficiency of the Google's Edge TPU and the Intel's Movidius Neural Compute Stick (NCS) for 3D-object detection. Hui et al [6] compared the accuracy in 3D object detection of different ML hardware acceleration platforms, i.e.…”

Section: Related Workmentioning

confidence: 99%

Exploring Deep Neural Networks on Edge TPU

Hosseininoorbin¹,

Layeghy²,

Kusý³

et al. 2021

Preprint

View full text Add to dashboard Cite

This paper explores the performance of Google's Edge TPU on feed forward neural networks. We consider Edge TPU as a hardware platform and explore different architectures of deep neural network classifiers, which traditionally has been a challenge to run on resource constrained edge devices. Based on the use of a joint-time-frequency data representation, also known as spectrogram, we explore the trade-off between classification performance and the energy consumed for inference. The energy efficiency of Edge TPU is compared with that of widely-used embedded CPU ARM Cortex-A53. Our results quantify the impact of neural network architectural specifications on the Edge TPU's performance, guiding decisions on the TPU's optimal operating point, where it can provide high classification accuracy with minimal energy consumption. Also, our evaluations highlight the crossover in performance between the Edge TPU and Cortex-A53, depending on the neural network specifications. Based on our analysis, we provide a decision chart to guide decisions on platform selection based on the model parameters and context.

show abstract

“…The Edge TPU platform is a purpose-built ASIC hardware accelerator recently developed by Google to run inference at the edge [1]. The hardware accelerators are created to enhance the computational capabilities of the systems and make it possible to implement DL algorithms on the edge devices [2,3,4]. The NIDS algorithms utilised in this exploration are selected from two major deep neural network architectures, the feed forward and convolutional neural networks.…”

Section: Introductionmentioning

confidence: 99%

Exploring Edge TPU for Network Intrusion Detection in IoT

Hosseininoorbin,

Layeghy,

Sarhan

et al. 2021

Preprint

View full text Add to dashboard Cite

This paper explores Google's Edge TPU for implementing a practical network intrusion detection system (NIDS) at the edge of IoT, based on a deep learning approach. While there are a significant number of related works that explore machine learning based NIDS for the IoT edge, they generally do not consider the issue of the required computational and energy resources. The focus of this paper is the exploration of deep learning-based NIDS at the edge of IoT, and in particular the computational and energy efficiency. In particular, the paper studies Google's Edge TPU as a hardware platform, and considers the following three key metrics: computation (inference) time, energy efficiency and the traffic classification performance. Various scaled model sizes of two major deep neural network architectures are used to investigate these three metrics. The performance of the Edge TPU-based implementation is compared with that of an energy efficient embedded CPU (ARM Cortex A53). Our experimental evaluation shows some unexpected results, such as the fact that the CPU significantly outperforms the Edge TPU for small model sizes.

show abstract

Artificial Vision on Edge IoT Devices: A Practical Case for 3D Data Classification

Cited by 7 publications

References 14 publications

Characterization of Metallic Gold Nanoparticles in a Colloidal State by Artificial Vision

Characterization of Metallic Gold Nanoparticles in a Colloidal State by Artificial Vision

Exploring Deep Neural Networks on Edge TPU

Exploring Edge TPU for Network Intrusion Detection in IoT

Contact Info

Product

Resources

About