Physical structure recognition on ICDAR P 2013
99.3PrecisionNCGM
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| NCGMTrain Dataset=Sci. + IC13-P, Setup=Setup-B2021.11 | 99.3 | 99.9 | 99.6 | |
| FLAG-NetTrain Dataset=Sci. + IC13-P, Setup=Setup-B2021.11 | 99.2 | 99.5 | 99.3 | |
| TabStruct-NetTrain Dataset=Sci. + IC13-P, Setup=Setup-B2021.11 | 99.1 | 99.3 | 99.2 | |
| DGCNNTrain Dataset=Sci. + IC13-P, Setup=Setup-B2021.11 | 98.6 | 99 | 98.8 | |
| NCGMTrain Dataset=Sci. + IC13-P, Setup=Setup-A2021.11 | 98.4 | 99.3 | 98.8 | |
| FLAG-NetTrain Dataset=Sci. + IC13-P, Setup=Setup-A2021.11 | 97.9 | 99.3 | 98.6 | |
| LGPMATrain Dataset=Sci. + IC13-P, Setup=Setup-A2021.11 | 96.7 | 99.1 | 97.9 | |
| Cycle-CenterNetTrain Dataset=WTW + IC19, Setup=Setup-A2021.11 | 95.5 | 88.3 | 91.7 | |
| GTETrain Dataset=Pub. + IC13-P, Setup=Setup-A2021.11 | 94.4 | 92.7 | 93.5 | |
| TabStruct-NetTrain Dataset=Sci. + IC13-P, Setup=Setup-A2021.11 | 93 | 90.8 | 91.9 |