Leaf Ying

icon location Hsinchu, Taiwan icon location Master student @CS-EVA-LAB. Member @COSCUP 2021 IT.

Results 7 comments of Leaf Ying

Question about the description of datasets in paper and the annotations in the repo

Hi, thanks for fast response. Sorry for describing my question not very clear. Due I cannot reproduce the annotation json file by the conversion script (because it doesn't set random...

[Feature] Support calculating loss in the validation step

Hi, I checked the current #1503 work. Except for the `ValLoop` changes, I have some suggestions about how to return the losses. # MMEngine To be more general and follow...

v2.7.1 pypi package - the `from blenderproc.scripts.visHdf5Files import vis_data` in `GifWriterUtility.py` is commented out, but the source code is inconsistent.

Hi, thanks for your fixing. After upgrading to 2.7.2, I can confirm the issue is fixed. But in the v2.7.1, it still exists. Because it is more like a packaging...

v2.7.1 pypi package - the `from blenderproc.scripts.visHdf5Files import vis_data` in `GifWriterUtility.py` is commented out, but the source code is inconsistent.

Hi Yes, the issue is resolved. Thanks!

Get a blurry result if I keep the same inference resolution as the original image (means scale = 1.0).

Hi, thanks for your response. Oh, I see. So, the actual problem may come from the re-shading network. But you say sending the image through as large as possible will...

Question about the benchmark results of the denoising methods

Hi, thanks for your quick reply. I understand the model performs better on noisy images than on denoised images. > I have not checked the object detection inference using the...

Bug: Incorrect dtype if training with the AMP after loss calculation

If I don't solve it, it will have chance to raise an exception, `RuntimeError: Found dtype Float but expected Half`, during `scaler.scale(loss).backward()`.