LocalFace: Learning significant local features for deep face recognition

Ke, Xiao; Lin, BingHui; Guo, Wenzhong

doi:10.1016/j.imavis.2022.104484

Cited by 5 publications

(2 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Computer vision technology has made it possible to use deep learning and image processing methods for biometric identification, including face recognition 7 , 8 . Many face recognition models and methods have been developed, such as DeepFace 9 , SphereFace 10 , central loss 11 , state-of-the-art face recognition models 12 , and LocalFace 13 . Similar methods have also been used in animal feature recognition tasks, such as automatic identification of individual cows 14 and goats 15 , pig face recognition 16 , cow face recognition 17 , 18 , and individual egg identification 19 .…”

Section: Introductionmentioning

confidence: 99%

A novel method for Pu-erh tea face traceability identification based on improved MobileNetV3 and triplet loss

Zhang

Yang

Luo

et al. 2023

Sci Rep

View full text Add to dashboard Cite

Ensuring the traceability of Pu-erh tea products is crucial in the production and sale of tea, as it is a key means to ensure their quality and safety. The common approach used in traceability systems is the utilization of bound Quick Response (QR) codes or Near Field Communication (NFC) chips to track every link in the supply chain. However, counterfeiting risks still persist, as QR codes or NFC chips can be copied and inexpensive products can be fitted into the original packaging. To address this issue, this paper proposes a tea face verification model called TeaFaceNet for traceability verification. The aim of this model is to improve the traceability of Pu-erh tea products by quickly identifying counterfeit products and enhancing the credibility of Pu-erh tea. The proposed method utilizes an improved MobileNetV3 combined with Triplet Loss to verify the similarity between two input tea face images with different texture features. The recognition accuracy of the raw tea face dataset, ripe tea face dataset and mixed tea face dataset of the TeaFaceNet network were 97.58%, 98.08% and 98.20%, respectively. Accurate verification of tea face was achieved using the optimal threshold. In conclusion, the proposed TeaFaceNet model presents a promising approach to enhance the traceability of Pu-erh tea products and combat counterfeit products. The robustness and generalization ability of the model, as evidenced by the experimental results, highlight its potential for improving the accuracy of Pu-erh tea face recognition and enhancing the credibility of Pu-erh tea in the market. Further research in this area is warranted to advance the traceability of Pu-erh tea products and ensure their quality and safety.

show abstract

Section: Introductionmentioning

confidence: 99%

A novel method for Pu-erh tea face traceability identification based on improved MobileNetV3 and triplet loss

Zhang

Yang

Luo

et al. 2023

Sci Rep

View full text Add to dashboard Cite

show abstract

“…But generally, these models were originally designed specifically for certain objects, so the CNN model's performance can only work optimally on certain objects. For example, in research [27] for multiple object detection, research [28] for face recognition, [29] for face mask recognition, and [30] for violence detection. This proves that the design of the CNN model for handwriting recognition, especially Javanese script, is needed to get optimal recognition performance.…”

Section: Introductionmentioning

confidence: 99%

Improved Javanese script recognition using custom model of convolution neural network

Susanto,

Mulyono,

Sari

et al. 2023

IJECE

View full text Add to dashboard Cite

<div align="center"><span lang="EN-US">Handwriting recognition in Javanese script is not widely developed with deep learning (DL). Previous DL and machine learning (ML) research is generally limited to basic characters (Carakan) only. This study proposes a deep learning model using a custom-built convolutional neural network to improve recognition accuracy performance and reduce computational costs. The main features of handwritten objects are textures, edges, lines, and shapes, so convolution layers are not designed in large numbers. This research maximizes optimization of other layers such as pooling, activation function, fully connected layer, optimizer, and parameter settings such as dropout and learning rate. There are eleven main layers used in the proposed custom convolutional neural network (CNN) model, namely four convolution layers+activation function, four pooling layers, two fully connected layers, and a softmax classifier. Based on the test results on the Javanese script handwritten image dataset with 120 classes consisting of 20 basic character classes and 100 compound character classes, the resulting accuracy is 97.29%.</span></div>

show abstract

Reconstruction with robustness: A semantic prior guided face super-resolution framework for multiple degradations

Wu,

Qi,

Zhang

et al. 2023

Image and Vision Computing

View full text Add to dashboard Cite

LocalFace: Learning significant local features for deep face recognition

Cited by 5 publications

References 1 publication

A novel method for Pu-erh tea face traceability identification based on improved MobileNetV3 and triplet loss

A novel method for Pu-erh tea face traceability identification based on improved MobileNetV3 and triplet loss

Improved Javanese script recognition using custom model of convolution neural network

Reconstruction with robustness: A semantic prior guided face super-resolution framework for multiple degradations

Contact Info

Product

Resources

About