Semantic Entity Labeling on FUNSD v1 (test)
83.24PrecisionLayoutLMv2-large
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| LayoutLMv2-large#param (M)=426, Input Modalities=Image + Text + Spatial2021.06 | 83.24 | 85.19 | 84.2 | |
| DocFormer-large (I+T+S)#param (M)=536, Input Modalities=Image + Text + Spatial2021.06 | 82.29 | 86.94 | 84.55 | |
| DocFormer-large (T+S)#param (M)=536, Input Modalities=Text + Spatial2021.06 | 81.33 | 85.44 | 83.33 | |
| DocFormer-base (I+T+S)#param (M)=183, Input Modalities=Image + Text + Spatial2021.06 | 80.76 | 86.09 | 83.34 | |
| BROS-base#param (M)=139, Input Modalities=Text + Spatial2021.06 | 80.56 | 81.88 | 81.21 | |
| LayoutLMv2-base#param (M)=200, Input Modalities=Image + Text + Spatial2021.06 | 80.29 | 85.39 | 82.76 | |
| DocFormer-base (T+S)#param (M)=149, Input Modalities=Text + Spatial2021.06 | 77.63 | 83.69 | 80.54 | |
| LayoutLMv1-base#param (M)=160, Input Modalities=Image + Text + Spatial2021.06 | 76.77 | 81.95 | 79.27 | |
| LayoutLMv1-base#param (M)=113, Input Modalities=Text + Spatial2021.06 | 76.12 | 81.55 | 78.66 | |
| LayoutLMv1-large#param (M)=343, Input Modalities=Text + Spatial2021.06 | 75.36 | 80.61 | 77.89 | |
| ROBERTa-large#param (M)=355, Input Modalities=Text + Spatial2021.06 | 67.8 | 73.91 | 70.72 | |
| UniLMv2-large#param (M)=355, Input Modalities=Text + Spatial2021.06 | 67.8 | 73.91 | 70.72 | |
| ROBERTa-base#param (M)=125, Input Modalities=Text + Spatial2021.06 | 63.49 | 69.75 | 66.48 | |
| UniLMv2-base#param (M)=125, Input Modalities=Text + Spatial2021.06 | 63.49 | 69.75 | 66.48 | |
| BERT-large#param (M)=340, Input Modalities=Text + Spatial2021.06 | 61.13 | 70.85 | 65.63 | |
| BERT-base#param (M)=109, Input Modalities=Text + Spatial2021.06 | 54.69 | 61.71 | 60.26 |