Sign language recognition is vital for enhancing communication accessibility among the Deaf and hard-of-hearing communities.In Japan,approximately 360,000 individualswith hearing and speech disabilities rely on Japane...Sign language recognition is vital for enhancing communication accessibility among the Deaf and hard-of-hearing communities.In Japan,approximately 360,000 individualswith hearing and speech disabilities rely on Japanese Sign Language(JSL)for communication.However,existing JSL recognition systems have faced significant performance limitations due to inherent complexities.In response to these challenges,we present a novel JSL recognition system that employs a strategic fusion approach,combining joint skeleton-based handcrafted features and pixel-based deep learning features.Our system incorporates two distinct streams:the first stream extracts crucial handcrafted features,emphasizing the capture of hand and body movements within JSL gestures.Simultaneously,a deep learning-based transfer learning stream captures hierarchical representations of JSL gestures in the second stream.Then,we concatenated the critical information of the first stream and the hierarchy of the second stream features to produce the multiple levels of the fusion features,aiming to create a comprehensive representation of the JSL gestures.After reducing the dimensionality of the feature,a feature selection approach and a kernel-based support vector machine(SVM)were used for the classification.To assess the effectiveness of our approach,we conducted extensive experiments on our Lab JSL dataset and a publicly available Arabic sign language(ArSL)dataset.Our results unequivocally demonstrate that our fusion approach significantly enhances JSL recognition accuracy and robustness compared to individual feature sets or traditional recognition methods.展开更多
Sign language,a visual-gestural language used by the deaf and hard-of-hearing community,plays a crucial role in facilitating communication and promoting inclusivity.Sign language recognition(SLR),the process of automa...Sign language,a visual-gestural language used by the deaf and hard-of-hearing community,plays a crucial role in facilitating communication and promoting inclusivity.Sign language recognition(SLR),the process of automatically recognizing and interpreting sign language gestures,has gained significant attention in recent years due to its potential to bridge the communication gap between the hearing impaired and the hearing world.The emergence and continuous development of deep learning techniques have provided inspiration and momentum for advancing SLR.This paper presents a comprehensive and up-to-date analysis of the advancements,challenges,and opportunities in deep learning-based sign language recognition,focusing on the past five years of research.We explore various aspects of SLR,including sign data acquisition technologies,sign language datasets,evaluation methods,and different types of neural networks.Convolutional Neural Networks(CNN)and Recurrent Neural Networks(RNN)have shown promising results in fingerspelling and isolated sign recognition.However,the continuous nature of sign language poses challenges,leading to the exploration of advanced neural network models such as the Transformer model for continuous sign language recognition(CSLR).Despite significant advancements,several challenges remain in the field of SLR.These challenges include expanding sign language datasets,achieving user independence in recognition systems,exploring different input modalities,effectively fusing features,modeling co-articulation,and improving semantic and syntactic understanding.Additionally,developing lightweight network architectures for mobile applications is crucial for practical implementation.By addressing these challenges,we can further advance the field of deep learning for sign language recognition and improve communication for the hearing-impaired community.展开更多
This paper addresses the application of hand gesture recognition in monocular image sequences using Active Appearance Model (AAM), For this work, the proposed algorithm is composed of constricting AAMs and fitting t...This paper addresses the application of hand gesture recognition in monocular image sequences using Active Appearance Model (AAM), For this work, the proposed algorithm is composed of constricting AAMs and fitting the models to the interest region. In training stage, according to the manual labeled feature points, the relative AAM is constructed and the corresponding average feature is obtained. In recognition stage, the interesting hand gesture region is firstly segmented by skin and movement cues. Secondly, the models are fitted to the image that includes the hand gesture, and the relative features are extracted. Thirdly, the classification is done by comparing the extracted features and average features. 30 different gestures of Chinese sign language are applied for testing the effectiveness of the method. The Experimental results are given indicating good performance of the algorithm.展开更多
In a general overview, signed language is a technique used for communicational purposes by deaf people. It is a three-dimensional language that relies on visual gestures and moving hand signs that classify letters and...In a general overview, signed language is a technique used for communicational purposes by deaf people. It is a three-dimensional language that relies on visual gestures and moving hand signs that classify letters and words. Gesture recognition has been always a relatively fearful subject that is adherent to the individual on both academic and demonstrative levels. The core objective of this system is to produce a method which can identify detailed humanoid nods and use them to either deliver ones thoughts and feelings, or for device control. This system will stand as an effective replacement for speech, enhancing the individual’s ability to express and intermingle in society. In this paper, we will discuss the different steps used to input, recognize and analyze the hand gestures, transforming them to both written words and audible speech. Each step is an independent algorithm that has its unique variables and conditions.展开更多
Hand gesture recognition system is considered as a way for more intuitive and proficient human computer interaction tool. The range of applications includes virtual prototyping, sign language analysis and medical trai...Hand gesture recognition system is considered as a way for more intuitive and proficient human computer interaction tool. The range of applications includes virtual prototyping, sign language analysis and medical training. In this paper, an efficient Indian Sign Language Recognition System (ISLR) is proposed for deaf and dump people using hand gesture images. The proposed ISLR system is considered as a pattern recognition technique that has two important modules: feature extraction and classification. The joint use of Discrete Wavelet Transform (DWT) based feature extraction and nearest neighbour classifier is used to recognize the sign language. The experimental results show that the proposed hand gesture recognition system achieves maximum 99.23% classification accuracy while using cosine distance classifier.展开更多
A multidisciplinary approach for developing an intelligent sign multi-language recognition system to greatly enhance deaf-mute communication will be discussed and implemented. This involves designing a low-cost glove-...A multidisciplinary approach for developing an intelligent sign multi-language recognition system to greatly enhance deaf-mute communication will be discussed and implemented. This involves designing a low-cost glove-based sensing system, collecting large and diverse datasets, preprocessing the data, and using efficient machine learning models. Furthermore, the glove is integrated with a user-friendly mobile application called “Life-sign” for this system. The main goal of this work is to minimize the processing time of machine learning classifiers while maintaining higher accuracy performance. This is achieved by using effective preprocessing algorithms to handle noisy and inconsistent data. Testing and iterating approaches have been applied to various classifiers to refine and improve their accuracy in the recognition process. Additionally, the Extra Trees (ET) classifier has been identified as the best algorithm, with results proving successful gesture prediction at an average accuracy of about 99.54%. A smart optimization feature has been implemented to control the size of data transferred via Bluetooth, allowing for fast recognition of consecutive gestures. Real-time performance has been measured through extensive experimental testing on various consecutive gestures, specifically referring to Arabic Sign Language (ArSL). The results have demonstrated that the system guarantees consecutive gesture recognition with a lower delay of 50 milliseconds.展开更多
基金supported by the Competitive Research Fund of the University of Aizu,Japan.
文摘Sign language recognition is vital for enhancing communication accessibility among the Deaf and hard-of-hearing communities.In Japan,approximately 360,000 individualswith hearing and speech disabilities rely on Japanese Sign Language(JSL)for communication.However,existing JSL recognition systems have faced significant performance limitations due to inherent complexities.In response to these challenges,we present a novel JSL recognition system that employs a strategic fusion approach,combining joint skeleton-based handcrafted features and pixel-based deep learning features.Our system incorporates two distinct streams:the first stream extracts crucial handcrafted features,emphasizing the capture of hand and body movements within JSL gestures.Simultaneously,a deep learning-based transfer learning stream captures hierarchical representations of JSL gestures in the second stream.Then,we concatenated the critical information of the first stream and the hierarchy of the second stream features to produce the multiple levels of the fusion features,aiming to create a comprehensive representation of the JSL gestures.After reducing the dimensionality of the feature,a feature selection approach and a kernel-based support vector machine(SVM)were used for the classification.To assess the effectiveness of our approach,we conducted extensive experiments on our Lab JSL dataset and a publicly available Arabic sign language(ArSL)dataset.Our results unequivocally demonstrate that our fusion approach significantly enhances JSL recognition accuracy and robustness compared to individual feature sets or traditional recognition methods.
基金supported from the National Philosophy and Social Sciences Foundation(Grant No.20BTQ065).
文摘Sign language,a visual-gestural language used by the deaf and hard-of-hearing community,plays a crucial role in facilitating communication and promoting inclusivity.Sign language recognition(SLR),the process of automatically recognizing and interpreting sign language gestures,has gained significant attention in recent years due to its potential to bridge the communication gap between the hearing impaired and the hearing world.The emergence and continuous development of deep learning techniques have provided inspiration and momentum for advancing SLR.This paper presents a comprehensive and up-to-date analysis of the advancements,challenges,and opportunities in deep learning-based sign language recognition,focusing on the past five years of research.We explore various aspects of SLR,including sign data acquisition technologies,sign language datasets,evaluation methods,and different types of neural networks.Convolutional Neural Networks(CNN)and Recurrent Neural Networks(RNN)have shown promising results in fingerspelling and isolated sign recognition.However,the continuous nature of sign language poses challenges,leading to the exploration of advanced neural network models such as the Transformer model for continuous sign language recognition(CSLR).Despite significant advancements,several challenges remain in the field of SLR.These challenges include expanding sign language datasets,achieving user independence in recognition systems,exploring different input modalities,effectively fusing features,modeling co-articulation,and improving semantic and syntactic understanding.Additionally,developing lightweight network architectures for mobile applications is crucial for practical implementation.By addressing these challenges,we can further advance the field of deep learning for sign language recognition and improve communication for the hearing-impaired community.
文摘This paper addresses the application of hand gesture recognition in monocular image sequences using Active Appearance Model (AAM), For this work, the proposed algorithm is composed of constricting AAMs and fitting the models to the interest region. In training stage, according to the manual labeled feature points, the relative AAM is constructed and the corresponding average feature is obtained. In recognition stage, the interesting hand gesture region is firstly segmented by skin and movement cues. Secondly, the models are fitted to the image that includes the hand gesture, and the relative features are extracted. Thirdly, the classification is done by comparing the extracted features and average features. 30 different gestures of Chinese sign language are applied for testing the effectiveness of the method. The Experimental results are given indicating good performance of the algorithm.
文摘In a general overview, signed language is a technique used for communicational purposes by deaf people. It is a three-dimensional language that relies on visual gestures and moving hand signs that classify letters and words. Gesture recognition has been always a relatively fearful subject that is adherent to the individual on both academic and demonstrative levels. The core objective of this system is to produce a method which can identify detailed humanoid nods and use them to either deliver ones thoughts and feelings, or for device control. This system will stand as an effective replacement for speech, enhancing the individual’s ability to express and intermingle in society. In this paper, we will discuss the different steps used to input, recognize and analyze the hand gestures, transforming them to both written words and audible speech. Each step is an independent algorithm that has its unique variables and conditions.
文摘Hand gesture recognition system is considered as a way for more intuitive and proficient human computer interaction tool. The range of applications includes virtual prototyping, sign language analysis and medical training. In this paper, an efficient Indian Sign Language Recognition System (ISLR) is proposed for deaf and dump people using hand gesture images. The proposed ISLR system is considered as a pattern recognition technique that has two important modules: feature extraction and classification. The joint use of Discrete Wavelet Transform (DWT) based feature extraction and nearest neighbour classifier is used to recognize the sign language. The experimental results show that the proposed hand gesture recognition system achieves maximum 99.23% classification accuracy while using cosine distance classifier.
文摘A multidisciplinary approach for developing an intelligent sign multi-language recognition system to greatly enhance deaf-mute communication will be discussed and implemented. This involves designing a low-cost glove-based sensing system, collecting large and diverse datasets, preprocessing the data, and using efficient machine learning models. Furthermore, the glove is integrated with a user-friendly mobile application called “Life-sign” for this system. The main goal of this work is to minimize the processing time of machine learning classifiers while maintaining higher accuracy performance. This is achieved by using effective preprocessing algorithms to handle noisy and inconsistent data. Testing and iterating approaches have been applied to various classifiers to refine and improve their accuracy in the recognition process. Additionally, the Extra Trees (ET) classifier has been identified as the best algorithm, with results proving successful gesture prediction at an average accuracy of about 99.54%. A smart optimization feature has been implemented to control the size of data transferred via Bluetooth, allowing for fast recognition of consecutive gestures. Real-time performance has been measured through extensive experimental testing on various consecutive gestures, specifically referring to Arabic Sign Language (ArSL). The results have demonstrated that the system guarantees consecutive gesture recognition with a lower delay of 50 milliseconds.