method: baseline - RoBERTa-base with synthetic pre-training2023-05-02
Authors: Organizers
Affiliation: Rossum.ai, Czech Technical University in Prague, University of La Rochelle
Description: Baseline method. Uses multi-label NER formulation with RoBERTa base as the backbone. It is pre-trained on the synthetic part of the DocILE dataset.
method: baseline - LayoutLMv3 with unsupervised and synthetic pre-training2023-05-02
Authors: Organizers
Affiliation: Rossum.ai, Czech Technical University in Prague, University of La Rochelle
Description: Baseline method. Uses multi-label NER formulation with LayoutLMv3 as the backbone. It is pre-trained on the unlabelled and synthetic parts of the DocILE dataset.
method: baseline - RoBERTa-base2023-05-02
Authors: Organizers
Affiliation: Rossum.ai, Czech Technical University in Prague, University of La Rochelle
Description: Baseline method. Uses multi-label NER formulation with RoBERTa base as the backbone.
Description Paper Source Code
| Date | Method | F1 | AP | Precision | Recall | |||
|---|---|---|---|---|---|---|---|---|
| 2023-05-02 | baseline - RoBERTa-base with synthetic pre-training | 64.91% | 51.35% | 65.99% | 63.87% | |||
| 2023-05-02 | baseline - LayoutLMv3 with unsupervised and synthetic pre-training | 64.54% | 52.38% | 66.30% | 62.86% | |||
| 2023-05-02 | baseline - RoBERTa-base | 63.68% | 50.52% | 64.44% | 62.94% | |||
| 2023-05-02 | baseline - LayoutLMv3 with unsupervised pre-training | 61.98% | 47.92% | 63.96% | 60.13% | |||
| 2023-05-25 | GraphDoc+Classify+Merge | 61.78% | 38.32% | 65.76% | 58.25% | |||
| 2023-05-24 | YOLOv8X+Grid | 0.00% | 0.00% | 0.00% | 0.00% |