Table 5 Selected past research results on static gesture recognition.
From: Sign language recognition based on dual-path background erasure convolutional neural network
Specificities | Dataset | Accuracy (%) |
|---|---|---|
skeletonization algorithm + CNN42 | ASK gesture database | 96.01 |
Dual-path depth-aware attention network37 | ASL Finger spelling dataset | 93.53 |
Color moment + Hu moment + Gray Level Cooccurrence Matrix + SVM43 | American sign language | 87.00 |
media-pipe + SVM + GBM44 | ASL Finger spelling dataset | 98.45 |
CNN45 | HUST-ASL Dataset | 98.93 |
multi‑headed CNN36 | ASL Finger spelling dataset | 98.98 |
CNN46 | IPN Hand dataset | 87.5 |
Attention based graph47 | SHREC’17 | 97.01 |
DPCNN | ASL Finger spelling dataset | 99.52 |
