Research Article
Evaluation of Vision Transformers for Traffic Sign Classification
| Hyperparameter | Value |
| Number of classes | Refer to Table 1 | Image patch size | 32 | Image pixel size | 4 | Dimension of patch token | 256 | Dimension of pixel token | 12 | Number of TNT blocks | 4 | Attention dropout rate | 0.1 | Feedforward dropout rate | 0.1 |
|
|