askerlee

This is after conv_proj_q, conv_proj_k and conv_proj_v. But I'm not sure why the authors still use the pointwise projections after the conv projections.

@triangleCZH Thanks for pointing that out. We can also resize the input images to 64x64, and remove the first pooling layer. The final feature maps will also be 4x4, and...

@920232796 可否分享下pretrained model呢？多谢！

The sentence information is actually not used. So it should not impact the performance. Do you mean that dots are part of the abbreviations? In this case you could modify...

It supports corpus without categories. If you don't specify "-s" for topicExp.py, it does inference by assuming all documents sharing the same set of topic embeddings, i.e., no category information...

Seems the groundtruth labels are not correctly loaded?

Thanks for reporting the bug. I've just corrected "refuge" to "fundus". Also I've simplified the polyformer config. Yes you are right. Fine-tuning k is only to finetune the k of...

> “Also I've simplified the polyformer config.” so what file i should replace？ You can just do a "git pull origin master" to update the code. > in addition may...

> in addition, the poly dataset i downloaded has different folders... can you please upload your processed data？ You mean polyp? For people in China: https://pan.baidu.com/s/1TuiPyQirN4J2hQfQxMMHkQ?pwd=whjl For people in other...

mask_batch shape: torch.Size([1, 3, 320, 320]) This means your masks have 3 channels, i.e., three segmentation classes. I hard-coded the polyp task to have 2 classes (according to my dataset)....