An eight-layer convolutional neural network with stochastic pooling, batch normalization and dropout for fingerspelling recognition of Chinese sign language

Jiang, Xianwei; Lu, Mingzhou; Wang, Shuihua

doi:10.1007/s11042-019-08345-y

Cited by 33 publications

(20 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For example, batch normalization (BN) is introduced to ensure that the each layer input can be kept uniformly distributed during training. Dropout technology is employed to obtain a refined network with fewer nodes by dropping appropriate neurons based on probability P, thereby effectively reducing overfitting and achieving regularization to a certain extent [70]. ReLU function is focused on because it is easier to accelerate the convergence rate of stochastic gradient descent than traditional activation functions [71].…”

Section: Other Classification Approachesmentioning

confidence: 99%

“…ReLU function is focused on because it is easier to accelerate the convergence rate of stochastic gradient descent than traditional activation functions [71]. Data augmentation (DA) provides multiple augment methods, including PCA color enhancement, affine transformation, noise injection, gamma correction, random shifting, and scaling, which will effectively enlarge the dataset and be helpful to alleviate overfitting [70].…”

Section: Other Classification Approachesmentioning

confidence: 99%

“…As can be seen from Table 3, major Chinese Sign Language Recognition-related papers and their focus were listed, including conferences, journals, and workshops. These articles contain sensor-based and vision-based recognition methods, introducing some advanced and fashionable technologies such as SVM [23,24,39,40,101,102], DTW [103][104][105], HMM [20,22,23,101,[106][107][108][109][110][111][112][113][114], LSTM [115], ANN [116], CNN (3D-CNN) [70,71,117,118], HOD, HOG [23,24,103], and PCA [20]. Hidden Markov model (HMM) is a general processing method of Sign Language Recognition.…”

Section: Investigation Of Chinese Sign Language Recognitionmentioning

confidence: 99%

“…A private dataset including 40 vocabularies was evaluated to validate their approach and accuracy achieved 99%. Jiang et al [70,71] proposed two types of vision-based deep convolutional neural networks with advanced techniques, such [118]. They tested it on a private large-scale sign language dataset including 500 categories of CSL and on the ChalLearn14 benchmark and gained effective results.…”

Section: Investigation Of Chinese Sign Language Recognitionmentioning

confidence: 99%

“…As for classification approaches, HMM [20,22,[106][107][108][109][110][111][112][113][114] and SVM [23,24,39,40,101,102] are the most utilized methods. In recent years, the convolutional neural network [70,71,117,146] achieves more popularity in CSLR researches. As can be seen from Fig.…”

Section: Characteristics Of Chinese Sign Language Recognitionmentioning

confidence: 99%

See 4 more Smart Citations

A Survey on Artificial Intelligence in Chinese Sign Language Recognition

et al. 2020

Self Cite

View full text Add to dashboard Cite

Chinese Sign Language (CSL) offers the main means of communication for the hearing impaired in China. Sign Language Recognition (SLR) can shorten the distance between the hearing-impaired and healthy people and help them integrate into the society. Therefore, SLR has become the focus of sign language application research. Over the years, the continuous development of new technologies provides a source and motivation for SLR. This paper aims to cover the most recent approaches in Chinese Sign Language Recognition (CSLR). With a thorough review of superior methods from 2000 to 2019 in CSLR researches, various techniques and algorithms such as scale-invariant feature transform, histogram of oriented gradients, wavelet entropy, Hu moment invariant, Fourier descriptor, gray-level co-occurrence matrix, dynamic time warping, principal component analysis, autoencoder, hidden Markov model (HMM), support vector machine (SVM), random forest, skin color modeling method, k-NN, artificial neural network, convolutional neural network (CNN), and transfer learning are discussed in detail, which are based on several major stages, that is, data acquisition, preprocessing, feature extraction, and classification. CSLR was summarized from some aspect as follows: methods of classification and feature extraction, accuracy/performance evaluation, and sample size/datasets. The advantages and limitations of different CSLR approaches were compared. It was found that data acquisition is mainly through Kinect and camera, and the feature extraction focuses on hand's shape and spatiotemporal factors, but ignoring facial expressions. HMM and SVM are used most in the classification. CNN is becoming more and more popular, and a deep neural network-based recognition approach will be the future trend. However, due to the complexity of the contemporary Chinese language, CSLR generally has a lower accuracy than other SLR. It is necessary to establish an appropriate dataset to conduct comparable experiments. The issue of decreasing accuracy as the dataset increases needs to resolve. Overall, our study is hoped to give a comprehensive presentation for those people who are interested in CSLR and SLR and to further contribute to the future research.

show abstract

Section: Other Classification Approachesmentioning

confidence: 99%

Section: Other Classification Approachesmentioning

confidence: 99%