search

IIGROUP / MANIQA

[CVPRW 2022] MANIQA: Multi-dimension Attention Network for No-Reference Image Quality Assessment
Apache License 2.0
307 stars 36 forks source link

Access pretrained weights and test model with different image size #1

Closed kiashann closed 2 years ago

kiashann commented 2 years ago

Hi, how we van access pretrained weights. Is this possible to use image with different sizes for test the model? Do you prepare any google colab demo for project? @Shuweis

Shuweis commented 2 years ago

Hi, thank you for your attention. We will release the MANIQA weights several days later. For different image size, you can crop 224*224 with multiple times and average the score to get the last result. We will prepare the google colab demo in the future.

TianheWu commented 2 years ago

Hi, how we van access pretrained weights. Is this possible to use image with different sizes for test the model? Do you prepare any google colab demo for project? @Shuweis

Hi，Our code includes the timm library. When you train your code, the timm library automatically loads VIT's pretrained models.

kiashann commented 2 years ago

Hi, how we van access pretrained weights. Is this possible to use image with different sizes for test the model? Do you prepare any google colab demo for project? @Shuweis

Hi，Our code includes the timm library. When you train your code, the timm library automatically loads VIT's pretrained models.

I want to use the model for inference without training. Could you please share the weight of the model?

TianheWu commented 2 years ago

Hi, how we van access pretrained weights. Is this possible to use image with different sizes for test the model? Do you prepare any google colab demo for project? @Shuweis

Hi，Our code includes the timm library. When you train your code, the timm library automatically loads VIT's pretrained models.

I want to use the model for inference without training. Could you please share the weight of the model?

Hi, we have already released the checkpoint. You can use the model for inference PIPAL22 Validation. Link: https://github.com/IIGROUP/MANIQA/releases/tag/PIPAL22-VALID-CKPT

kiashann commented 2 years ago

I can't use this files you mentioned above for run inference file. How I can run inference file without training with pretrained weight? @TianheWu