Research Article
PF-ViT: Parallel and Fast Vision Transformer for Offline Handwritten Chinese Character Recognition
Table 5
The parameters and FLOPs of the best T-ViT model proposed and other models.
| | Model | #Params (M) | FLOPs (G) |
| | ResNet-101 [46] | 44.7 | 7.9 | | Swin-B [47] | 88 | 15.4 | | CrossViT-18 [35] | 43.3 | 9.03 | | T-ViT | 43.11 | 4.32 |
|
|
