Benchmark about Medmnist+

[Weixiang-Sun]

Can u release the benchmark of medmnist+ like the size 224*224?
best wish

[Weixiang-Sun]

https://zenodo.org/records/7782114
I find here, sry for disturb

[sdoerrich97]

Hey @Weixiang-Sun,

Exciting news! We've just rolled out the very first solid baseline performance evaluation for MedMNIST+, titled "Rethinking Model Prototyping through the MedMNIST+ Dataset Collection". This comprehensive evaluation covers 10 different deep learning models trained via 3 distinct training schemes across all 12 2D datasets and available image resolutions (28x28, 64x64, 128x128, and 224x224) within the MedMNIST+ dataset.

You can check out the paper here: Link to the paper
And dive into the source code on our GitHub repository: Link to the GitHub repository

We're excited about how this evaluation can establish a baseline for future research in the MedMNIST+ benchmark domain. Looking forward to hearing your thoughts and contributions!

Cheers,
Sebastian

[duducheng]

Hi @sdoerrich97 ,

It’s really an exciting update! Thanks for your contribution to the MedMNIST community! I will add a reference to your work in the MedMNIST readme.

Jiancheng