| Study | Year | Techniques | Evaluation metric |
| [153] | 2023 | DeepVision transformers | Accuracy, precision | [154] | 2023 | 8-Layer CNN | Accuracy | [155] | 2023 | KNN | Accuracy | [161] | 2023 | Attention-based Bi-LSTM | Accuracy | [150] | 2023 | Deep learning (DL) combined with CNN and RNN | Accuracy | [147] | 2023 | DNN | Accuracy with [email protected] | [146] | 2022 | CNN | Accuracy | [144] | 2022 | SVM | Accuracy | [143] | 2022 | Inaudible acoustic signal to estimate channel information and capture the sign language in real time | Accuracy | [162] | 2022 | Hybrid convolutional neural network + bidirectional long short-term memory (CNN + Bi-LSTM) | Peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), Fréchet inception distance (FID), temporal consistency metric (TCM) | [163] | 2022 | 3D convolution net | Accuracy | [156] | 2022 | CNN | Accuracy | [157] | 2022 | CNN, DCGAN | Accuracy | [164] | 2022 | VGG-19 | PSNR, SSIM, FID, TCM | [141] | 2021 | SVM | Accuracy, precision, recall, F1 score | [158] | 2021 | CNN | Accuracy | [159] | 2021 | 3DCNN | Accuracy | [160] | 2021 | CNN, RNN | Accuracy | [139] | 2021 | Spyder, TensorFlow, OpenCV, Keras | Accuracy | [140] | 2021 | KNN | Accuracy | [142] | 2021 | 2D CNN, SVD, and LSTM | Time recognition, accuracy | [125] | 2020 | 3D CNN Siamese network | Accuracy | [131] | 2020 | Conv3D | Sentence error rate (SER), accuracy | [122] | 2020 | ResNet-D model | Accuracy, time | [134] | 2020 | CNN | Accuracy, precision, recall, F1 score | [135] | 2020 | Hidden Markov model (HMM) | Accuracy | [123] | 2020 | CNN-LSTM-HMM | Accuracy | [136] | 2020 | CNN | Accuracy | [126] | 2020 | CNN | Accuracy | [130] | 2020 | Stochastic multistate (SMS) | WER | [124] | 2020 | CNN LSTM | Accuracy, precision, recall, F1 measure | [116] | 2019 | CNN | NA | [99] | 2019 | 3D-ResNet, CTC | WER | [97] | 2019 | Visual Geometry Group (VGG)-16, VGG-19 | Accuracy | [94] | 2019 | CNN | Accuracy | [93] | 2019 | Convolutional-based attention module (CBAM)-ResNet | Accuracy | [86] | 2019 | Neural network and QuadroConvPoolNet | Accuracy | [95] | 2019 | MLP, SVM, and CNN | Accuracy | [106] | 2019 | ANN, SVM, HMM | Accuracy | [117] | 2019 | CNN | Accuracy | [114] | 2019 | CNN, LSTM | Accuracy | [118] | 2019 | VGG-19 | Recognition rate | [100] | 2019 | K-means clustering | Accuracy | [96] | 2019 | Inception v3, MobileNet | Precision, recall, F1 score, accuracy | [107] | 2019 | LSTM | Accuracy | [108] | 2019 | ResNet50-BiLSTM, MobileNetV2-BiLSTM | Precision, recall, F1 score, accuracy | [98] | 2019 | Deep feedforward neural network | Accuracy | [113] | 2019 | CNN | Precision, recall, F1 score, accuracy | [88] | 2019 | WebGL, SiGML, CoreNLP | Recognition rate | [112] | 2019 | CNN | Accuracy | [92] | 2019 | 3DCNN | Accuracy | [89] | 2019 | CNN | Precision, recall, F1 score, accuracy | [71] | 2018 | SVM, KNN, CNN, ANN | Success rate | [72] | 2018 | LSTM and VGG-16 | Accuracy | [73] | 2018 | CNN | Accuracy | [77] | 2018 | CNN | Accuracy | [80] | 2018 | CNN | Accuracy | [85] | 2018 | Adaptive graph matching | Accuracy, TWRF, FWRF | [81] | 2018 | Restricted Boltzmann machine | Top-1 accuracy, Top-5 accuracy | [78] | 2018 | Inception v3 | Accuracy | [69] | 2018 | RNN | Accuracy | [76] | 2018 | LSTM | Accuracy | [68] | 2017 | Dynamic vision sensor, CNN, RNN | Accuracy | [64] | 2017 | 3D signing avatar, Blender animation software | Accuracy | [58] | 2017 | Nearest neighbor | Accuracy | [63] | 2017 | CNN | Accuracy | [59] | 2017 | Finite Legendre transform, linear discriminant analysis, KNN | Accuracy | [55] | 2017 | LSTM | Accuracy | [65] | 2017 | CNN | Accuracy | [48] | 2016 | CNN | Top-1 accuracy, Top-5 accuracy | [47] | 2016 | Hybrid-CNN HMM | Accuracy | [3] | 2016 | Correlation classification algorithm | Accuracy, precision, recall | [44] | 2016 | CNN | Accuracy | [46] | 2016 | SVM | Accuracy | [54] | 2016 | Maximum a posteriori (MAP) | Accuracy | [43] | 2015 | Leap Motion Technology | Accuracy | [36] | 2015 | CNN | Accuracy | [34] | 2014 | ANN, vision-based | Accuracy, MSE | [31] | 2014 | A skin and motion detector, hand detection using multiple proposals, chains model | Accuracy | [29] | 2014 | KNN, cross-correlation | Accuracy | [30] | 2014 | Deep belief network | Precision, recall | [32] | 2014 | Microcontroller | Precision, recall | [27] | 2013 | K-nearest neighbor | Accuracy | [25] | 2012 | Multilayer perceptron | Mean, standard deviation, aspect ratio hand cropping algorithm (ARHCA), no ARHCA |
|
|