Table 10 One example that is incorrectly extracted in both the test set from the literature and the test set generated by CDK
From: SwinOCSR: end-to-end optical chemical structure recognition using a Swin Transformer
Items | Molecule 1 |
|---|---|
The real-world image derived from the literature |
|
Manual-labeled SMILES | c1c(cc(c(c1[Y1])[X0])[Y2])c2c(cc([H][H][R0])cc2[Y4])[Y3] |
Predicted SMILES from the real-world image | C1CC(CCC1C2CCC(CC2)[Y])c4cc(c(-c3cc(c(c(c3)[Y1])[Y1])[Y])c(c4)[Y])[Y] |
Generated image from the above-mentioned predicted SMILES |
|
Generated image from manual-labeled SMILES by CDK |
|
Predicted SMILES from the generated image | c1c(cc(c(c1[Y1])[X0])[Y2])-c2c(cc(cc2[Y4])N[R0])[Y4] |
Generated image from the above-mentioned predicted SMILES |
|
