Update README.md

README.md

@@ -24,10 +24,30 @@ For semantic segmentation, we will use the F1-score to evaluate the classificati
 Generalized Intersection over Union (GIoU) and the accuracy of the predicted bounding box class. The evaluation is conducted on 500 pages from the PDF evaluation
 dataset of DocLayNet.
 
+|      Class     | f1-score (x100) | GIoU (x100) | accuracy (x100) |
+|:--------------:|:---------------:|:-----------:|:---------------:|
+|   Background   |      91.69      |      NA     |        NA       |
+|     Caption    |      41.17      |    29.84    |      21.43      |
+|    Footnote    |        0        |    24.42    |        0        |
+|     Formula    |      66.74      |    43.28    |      66.30      |
+|    List-item   |      55.41      |    42.39    |      85.63      |
+|   Page-footer  |      70.95      |    -9.69    |      70.80      |
+|   Page-header  |      53.28      |    37.41    |      66.85      |
+|     Picture    |      68.66      |    64.90    |      93.31      |
+| Section-header |      50.52      |    20.90    |      50.54      |
+|      Table     |      70.31      |    69.28    |      70.40      |
+|      Text      |      81.87      |    52.53    |      77.96      |
+|      Title     |        0        |    31.90    |        0        |
+
 ## Benchmark
 
 Now, let's compare the performance of this model with other models.
 
+|      Class                            | f1-score (x100) | GIoU (x100) | accuracy (x100) |
+|:-------------------------------------:|:---------------:|:-----------:|:---------------:|
+| cmarkea/detr-layout-detection         |      84.23      |    43.84    |                 |
+| [cmarkea/dit-base-layout-detection]() |      90.77      |    56.29    |      85.26      |
+
 ## Direct Use
 
 ```python