search

xiaoman-zhang / PMC-VQA

PMC-VQA is a large-scale medical visual question-answering dataset, which contains 227k VQA pairs of 149k images that cover various modalities or diseases.
MIT License
174 stars 11 forks source link

The accuracy on the dataset version-2 #11

Open toggle1995 opened 1 year ago

toggle1995 commented 1 year ago

Hi, thanks for you amazing work. I find you release an update dataset with only noncompound images. So can you tell us the accuracy on this new version-2? I think the accuracy reported in the your paper is evaluated on version-1?

xiaoman-zhang commented 1 year ago

For the model pre-trained on the version-2 dataset, the accuracy on the multiple-choice task is 0.587, and on the open-ended blanking task is 0.343.

ShawnHuang497 commented 6 months ago

For the model pre-trained on the version-2 dataset, the accuracy on the multiple-choice task is 0.587, and on the open-ended blanking task is 0.343.

Hi, could you tell us the model architecture details about these metrics like Table 3 in your paper?

AmeeraBawazir commented 5 months ago

For the model pre-trained on the version-2 dataset, the accuracy on the multiple-choice task is 0.587, and on the open-ended blanking task is 0.343.

Hey, this evaluation result is for which model variant, the MedVInt-TE or MedVInt-TD?