Text extraction from images using the traditional techniques of image collecting,and pattern recognition using machine learning consume time due to the amount of extracted features from the images.Deep Neural Networks...Text extraction from images using the traditional techniques of image collecting,and pattern recognition using machine learning consume time due to the amount of extracted features from the images.Deep Neural Networks introduce effective solutions to extract text features from images using a few techniques and the ability to train large datasets of images with significant results.This study proposes using Dual Maxpooling and concatenating convolution Neural Networks(CNN)layers with the activation functions Relu and the Optimized Leaky Relu(OLRelu).The proposed method works by dividing the word image into slices that contain characters.Then pass them to deep learning layers to extract feature maps and reform the predicted words.Bidirectional Short Memory(BiLSTM)layers extractmore compelling features and link the time sequence fromforward and backward directions during the training phase.The Connectionist Temporal Classification(CTC)function calcifies the training and validation loss rates.In addition to decoding the extracted feature to reform characters again and linking them according to their time sequence.The proposed model performance is evaluated using training and validation loss errors on the Mjsynth and Integrated Argument Mining Tasks(IAM)datasets.The result of IAM was 2.09%for the average loss errors with the proposed dualMaxpooling and OLRelu.In the Mjsynth dataset,the best validation loss rate shrunk to 2.2%by applying concatenating CNN layers,and Relu.展开更多
The text watermarking is a feasible method to protect the copyright from being copied and tampered. In this paper, a text zero-watermarking algorithm is proposed based on the connection between the Chinese characters ...The text watermarking is a feasible method to protect the copyright from being copied and tampered. In this paper, a text zero-watermarking algorithm is proposed based on the connection between the Chinese characters and the Chinese phonetic alphabets. According to the predefined interval threshold, the proposed algorithm extracts the characteristics of the text content by valuing on the basis of the custom of Chinese phonetic alphabets. After being chaotic transformed, the algorithm combines the text characteristics with the embedded watermarking information in the Chinese text. The experimental results show that the watermarking＇s capability of preventing tampering is up to 0.1%, which demonstrates the strong robustness and resistance to aggressive behavior of the algorithm.展开更多
基金supported this project under the Fundamental Research Grant Scheme(FRGS)FRGS/1/2019/ICT02/UKM/02/9 entitled“Convolution Neural Network Enhancement Based on Adaptive Convexity and Regularization Functions for Fake Video Analytics”.This grant was received by Prof.Assis.Dr.S.N.H.Sheikh Abdullah,https://www.ukm.my/spifper/research_news/instrumentfunds.
文摘Text extraction from images using the traditional techniques of image collecting,and pattern recognition using machine learning consume time due to the amount of extracted features from the images.Deep Neural Networks introduce effective solutions to extract text features from images using a few techniques and the ability to train large datasets of images with significant results.This study proposes using Dual Maxpooling and concatenating convolution Neural Networks(CNN)layers with the activation functions Relu and the Optimized Leaky Relu(OLRelu).The proposed method works by dividing the word image into slices that contain characters.Then pass them to deep learning layers to extract feature maps and reform the predicted words.Bidirectional Short Memory(BiLSTM)layers extractmore compelling features and link the time sequence fromforward and backward directions during the training phase.The Connectionist Temporal Classification(CTC)function calcifies the training and validation loss rates.In addition to decoding the extracted feature to reform characters again and linking them according to their time sequence.The proposed model performance is evaluated using training and validation loss errors on the Mjsynth and Integrated Argument Mining Tasks(IAM)datasets.The result of IAM was 2.09%for the average loss errors with the proposed dualMaxpooling and OLRelu.In the Mjsynth dataset,the best validation loss rate shrunk to 2.2%by applying concatenating CNN layers,and Relu.
基金Supported by the National Natural Science Foundation of China(91112003)Youth Foundation(31541311307)
文摘The text watermarking is a feasible method to protect the copyright from being copied and tampered. In this paper, a text zero-watermarking algorithm is proposed based on the connection between the Chinese characters and the Chinese phonetic alphabets. According to the predefined interval threshold, the proposed algorithm extracts the characteristics of the text content by valuing on the basis of the custom of Chinese phonetic alphabets. After being chaotic transformed, the algorithm combines the text characteristics with the embedded watermarking information in the Chinese text. The experimental results show that the watermarking＇s capability of preventing tampering is up to 0.1%, which demonstrates the strong robustness and resistance to aggressive behavior of the algorithm.
