Thomas Viehmann

Results 227 comments of Thomas Viehmann

Just a quick shout in between: Thank you for going through this diligently and reporting the errors you found!

I think the last two are actually OK or I could use a hint what you think is unclear here.

Oh, yes. Sorry, we should have been more diligent in explaining this. I will likely try to take a stab at updating the code to the latest PyTorch soonish.

Don't despair! Data is a mess, and so this is a bit closer to real world data than, say, imagenet. These are good questions (and see the end of chapter...

Yes, the real-world coordinates are always in the Patient coordinate system shown in fig 10.6 and converted as 10.7. CNN detection accuracy is indeed resolution dependent, this is why we...

I think the license (CC-BY 4.0) allows it... but I would not know about the cost.

I completely agree. Thank you for pointing this out! I'd probably use `nn.init.zeros_` to keep with the `nn.init` theme.

@dzhwinter @wangkuiyi Say, would you be able to shed some light on the mask (0 if I see this correctly) you are using? in most other venues, it seems to...

So I think I have successfully found the knobs to turn for not running into unavoidable test failures. :)

@pytorchbot merge -g