Evaluation of deep neural network architectures in the identification of bone fissures

Sarmiento, Fredy Hernán Martínez; Hernández, César Arranz; Martínez, Fernando

doi:10.12928/telkomnika.v18i2.14754

Cited by 6 publications

(6 citation statements)

References 20 publications

(20 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The improvement of technology and popularization of big data analysis have facilitated the prevalence of video surveillance systems constructed with artificial intelligence (AI) deep learning [1]- [6], [7]- [10]. Because deep-learning methods require only a small number of samples and can complete training and validation within a short period of time, they are used in the recognition of human activities, facial expressions, and voices.…”

Section: Introductionmentioning

confidence: 99%

Real-time microreaction recognition system

Wu¹,

Liu²,

Huang³

2023

IJRA

View full text Add to dashboard Cite

<span lang="EN-US">This study constructed a real-time microreaction recognition system that can give real-time assistance to investigators. Test results indicated that the number of frames per second (30 or 190); angle of the camera, namely the front view of the interviewee or left (+45°) or right (−45°) view; and image resolution (480 or 680 p) did not have major effects on the system’s recognition ability. However, when the camera was placed at a distance of 300 cm, recognition did not always succeed. Value changes were larger when the camera was placed at an elevation 45° than when it was placed directly in front of the person being interrogated. Within a specific distance, the recognition results of the proposed real-time microreaction recognition system concurred with the six reaction case videos. In practice, only the distance and height of the camera must be adjusted in the real-time microreaction recognition system.</span>

show abstract

Section: Introductionmentioning

confidence: 99%

Real-time microreaction recognition system

Wu¹,

Liu²,

Huang³

2023

IJRA

View full text Add to dashboard Cite

show abstract

“…We propose the use of a neuronal model based on convolutional networks trained explicitly as an AI tool for the rapid and low-cost detection of individuals with COVID-19 [24]- [29]. For the model, we selected the NASNet deep network by Google Brain, due to its high performance against architectures like Inception-v2, Inception-v3, Xception, ResNet, and Inception-ResNet-v2 [30]- [32]. The model was optimized for a dataset with X-ray images taken from patients who have tested positive for COVID-19 and healthy people.…”

Section: Introductionmentioning

confidence: 99%

Performance Evaluation of the NASNet Convolutional Network in the Automatic Identification of COVID-19

Sarmiento¹,

Martínez²,

Jacinto³

2020

Int. J. Adv. Sci. Eng. Inf. Technol.

Self Cite

View full text Add to dashboard Cite

This paper evaluates the performance of the Neural Architecture Search Network (NASNet) in the automatic detection of COVID-19 (Coronavirus Disease 2019) from chest x-ray images. COVID-19 is a disease caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) that produces in patients fever, cough, shortness of breath, muscle pain, sputum production, diarrhea, and even sore throat. The virus spreads through the air, and to date, is expanding as a global pandemic. There is no vaccine, and it is fatal to approximately 2-7% of the infected population. Among the clinical and paraclinical characteristics of infected patients, nodules have been identified in images of chest x-rays that can be visually identified, producing a simple, rapid, and generally available method of identification. However, the rapid spread of the disease means that there is a lack of specialized medical personnel capable of identifying it, which is why automated schemes are being developed. We propose the tuning of a NASNet-type convolutional model to automatically determine the initial state of a patient in the triage process or intervention protocol of health care centers. The neural network is trained with public images of cases positively identified as patients infected with the virus and patients in normal conditions without infection. Performance evaluation is also done with real images unknown to the neuronal model. As for performance metrics, we use the function of loss of cross-entropy (categorical cross-entropy), the accuracy (or success rate), and the MSE (Mean Squared Error). The tuned model was able to correctly classify the test images with an accuracy of 97%.

show abstract

“…The proposed system for handgun detection in real-time is based on convolutional neural network (CNN). The CNNs has been rapidly growing in computer vision area during the past few years [3][4][5][6][7]. In computer vision there are three object detection models have been analyzed (faster region-based convolutional TELKOMNIKA Telecommun Comput El Control  The detection of handguns from live-video in real time based on deep learning (Muhammed Ghazal) 3027 neural networks (FR-CNN), region based-fully convolutional networks (R-FCN) and single shot detector (SSD)) [8].…”

Section: Introductionmentioning

confidence: 99%

The detection of handguns from live-video in real-time based on deep learning

2020

View full text Add to dashboard Cite

Many people have been killed indiscriminately by the use of handguns in different countries. Terrorist acts, online fighting games and mentally disturbed people are considered the common reasons for these crimes. A realtime handguns detection surveillance system is built to overcome these bad acts, based on convolutional neural networks (CNNs). This method is focused on the detection of different weapons, such as (handgun and rifles). The identification of handguns from surveillance cameras and images requires monitoring by human supervisor, that can cause errors. To overcome this issue, the designed detection system sends an alert message to the supervisor when a weapon is detected. In the proposed detection system, a pre-trained deep learning model MobileNetV3-SSDLite is used to perform the handgun detection operation. This model has been selected because it is fast and accurate in infering to integrate network for detecting and classifying weapons in images. The experimental result using global handguns datasets of various weapons showed that the use of MobileNetV3 with SSDLite model both enhance the accuracy level in identifying the real time handguns detection.

show abstract

Evaluation of deep neural network architectures in the identification of bone fissures

Cited by 6 publications

References 20 publications

Real-time microreaction recognition system

Real-time microreaction recognition system

Performance Evaluation of the NASNet Convolutional Network in the Automatic Identification of COVID-19

The detection of handguns from live-video in real-time based on deep learning

Contact Info

Product

Resources

About