A novel face recognition model for fighting against human trafficking in surveillance videos and rescuing victims

Karpagam, M.; Jeyavathana, R. Beaulah; Chinnappan, Sathiya Kumar; Kanimozhi, K.; Sambath, M.

doi:10.1007/s00500-022-06931-1

Cited by 7 publications

(3 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Cahyani et al [22] applied adaptive histogram equalization to address light sensitivity in the Eigenface method, enhancing image contrast and boosting success rate. Karpagam et al [23] developed a face recognition model for human trafficking prevention using surveillance videos, applying a Gaussian filter for noise reduction. Hang Du et al [24] highlighted the importance of face alignment in deep face recognition systems.…”

Section: Introductionmentioning

confidence: 99%

Evaluation of pseudorandom pixel placement in enhancing face recognition using VGG-Face

Hadis,

Akita,

Toda

2024

Fifth International Conference on Image, Video Processing, and Artificial Intelligence (IVPAI 2023)

View full text Add to dashboard Cite

In efforts to enhance face recognition performance, techniques ranging from super-resolution methods to the use of Local Binary Pattern (LBP) and deep learning have been explored. Among these, the pseudorandom pixel placement (PSE) technique has demonstrated potential in face recognition. Nevertheless, its testing was previously limited to just 8 subjects. This study undertakes a comprehensive evaluation of the PSE technique with a larger sample, utilizing 2000 subjects from the DigiFACE1M dataset and leveraging the state-of-the-art VGG-Face deep learning model. Through experiments involving 10 different PSE patterns on 144.000 face images, our findings indicate that, compared to Regular Pixel Placement (REG), PSE achieved an improvement in average accuracy by 1.05%, reduced the standard deviation by 1.47%, and resulted in 31 additional subjects achieving 100% accuracy. We conclude that PSE consistently outperforms REG in face recognition tasks using the VGG-Face model across the majority of tested scenarios.

show abstract

Section: Introductionmentioning

confidence: 99%

Evaluation of pseudorandom pixel placement in enhancing face recognition using VGG-Face

Hadis,

Akita,

Toda

2024

Fifth International Conference on Image, Video Processing, and Artificial Intelligence (IVPAI 2023)

View full text Add to dashboard Cite

show abstract

“…Deep learning-based methods, especially Convolutional Neural Networks (CNNs), have emerged as a new and effective solution to solve many computer vision-related problems, including agriculture [2,3], surveillance [4], healthcare [5,6], wildfire monitoring [7], and handwritten recognition [8]. In the last few years, several studies have targeted hot-rolled steel strip surface defect classification using different CNN architectures and training strategies.…”

Section: Introductionmentioning

confidence: 99%

CNN-based hot-rolled steel strip surface defects classification: a comparative study between different pre-trained CNN models

Bouguettaya,

Zarzour

2024

Int J Adv Manuf Technol

View full text Add to dashboard Cite

During the manufacturing process, hot-rolled steel strip surface defects occur frequently. These defects cause economic losses and risks in the use of these products. Therefore, it is crucial to develop automatic inspection systems to identify these defects. In the last few years, computer vision has emerged as an effective tool to identify these defects. Deep learning-based computer vision techniques, especially Convolutional Neural Networks (CNN), achieved state-of-the-art results for most computer vision tasks, including image classification. These results are obtained using a large amount of data. However, collecting such large datasets in the manufacturing field remains a challenging task. To overcome such a problem, a transfer learning-based framework with multiple CNN variants is proposed in this study. Therefore, different state-of-theart and widely used pre-trained CNN architectures, including VGG-16, VGG-19, ResNet50, ResNet50V2, InceptionV3, InceptionResNet-V2, MobileNet-V1, MobileNet-V2, MobileNet-V3 Small, and NASNetMobile, combined with transfer learning were investigated to evaluated their performances in classifying hot-rolled steel strips surface defects. The experimental results showed that MobileNet-V2 and InceptionResNet-V2-based methods achieve better performance than all other models in terms of accuracy, loss, training and inference times, and model size.

show abstract

Section: Introductionmentioning

confidence: 99%

CNN-based Hot-Rolled Steel Strip Surface Defects Classification: A Comparative Study Between Different Pre-Trained CNN Models

Bouguettaya¹,

Zarzour

2023

Preprint

View full text Add to dashboard Cite

During the manufacturing process, hot-rolled steel strip surface defects occur frequently. These defects cause economic losses and risks in the use of these products. Therefore, it is crucial to develop automatic inspection systems to identify these defects. In the last few years, computer vision has emerged as an effective tool to identify these defects. Deep learning-based computer vision techniques, especially Convolutional Neural Networks (CNN), achieved state-of-the-art results for most computer vision tasks, including image classification. These results are obtained using a large amount of data. However, collecting such large datasets in the manufacturing field remains a challenging task. To overcome such a problem, a transfer learning-based framework with multiple CNN variants is proposed in this study. Therefore, different state-of-theart and widely used pre-trained CNN architectures, including VGG-16, VGG-19, ResNet50, ResNet50V2, InceptionV3, InceptionResNet-V2, MobileNet-V1, MobileNet-V2, MobileNet-V3 Small, and NASNetMobile, combined with transfer learning were investigated to evaluated their performances in classifying hot-rolled steel strips surface defects. The experimental results showed that MobileNet-V2 and InceptionResNetV2- based methods achieve better performance than all other models in terms of accuracy, loss, training and inference times, and model size.

show abstract

A novel face recognition model for fighting against human trafficking in surveillance videos and rescuing victims

Cited by 7 publications

References 31 publications

Evaluation of pseudorandom pixel placement in enhancing face recognition using VGG-Face

Evaluation of pseudorandom pixel placement in enhancing face recognition using VGG-Face

CNN-based hot-rolled steel strip surface defects classification: a comparative study between different pre-trained CNN models

CNN-based Hot-Rolled Steel Strip Surface Defects Classification: A Comparative Study Between Different Pre-Trained CNN Models

Contact Info

Product

Resources

About