Abstract
As with the huge number of deaf-mute people in China is of concern, there is a growing need to integrate them into mainstream society through the use of efficient sign language processing technologies. Sign language processing entails the systematic recognition and translation of sign language images/videos to text or speech. This survey provides an overview of the most important work on Chinese sign language recognition and translation, discussed its classification, highlights the features explored in sign language recognition research, presents the datasets available, and provides trends for the future research.
Highlights
Sign language, as important as spoken language, is a visual gesture language among the deaf and hearing-impaired people
The model achieved 89.2% accuracy on their created Chinese sign language video in museum (SLVM) dataset (6800 samples of 20 vocabularies) and 92.4% on the Chalearn dataset
FOR FUTURE RESEARCH Chinese Sign Language Recognition system (CSLR) is an on-going research that began decades ago, but till there is no system deployed on a large scale
Summary
As important as spoken language, is a visual gesture language among the deaf and hearing-impaired people. The experimental results from their self-built dataset demonstrated the effectiveness of the proposed method that results in an accuracy of 93.5% and 94.7% for the image model and the fused model In another similar work, [28] proposed a novel Chinese Isolated sign language recognition approach based on Keyframe-Centered clips (KCC). Experiments were carried out on a self-built Kinect CSL words dataset, an accuracy of 89.87% and 91.18% is reported for different methods without KCC and with KCC, and proves to be better than HMM, DTW, CNNs, and LSTMs. [29] presented a 3D Convolutional Neural Networks for Sign Language Recognition. There exist several works on this type of SL, with the majority employing vision-based approaches with encouraging performances Further research in this category should be geared towards improving signer-independent word recognition and building an extensive dataset that contains all the current CSL words. CNNs [25], [29], LSTMs [18], [26], [34] or hybrid models [27], [33] have been adopted for isolated words and especially for continuous sentence recognition
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
