Matthijs Hollemans comments

Results 233 comments of


                                            Matthijs Hollemans

Getting error - 'BlazeFace' object has no attribute 'load_weights'

It works for me. What version of PyTorch are you using?

Update for SwiftUI

There is nothing in this book that is related to Swift UI. Not sure what there is to update?

Update for SwiftUI

I see your point but it's a little premature to call UIKit "legacy". I don't think I'm going to update the code samples to SwiftUI anytime soon, but feel free...

Update for SwiftUI

Sorry, but I currently don't offer discount codes.

the front and back has the same anchors?

I don't know. This was a contribution and I didn't test it myself.

Question about performance of SSDLite with Core ML

I didn't actually do any speed tests (hehe) but 2x speed difference is quite a bit! I wouldn't have expected that... Doing these operations inside the model will run them...

bit exact?

It should give exactly the same results. I used it with pretty good success in this competition: https://www.kaggle.com/c/deepfake-detection-challenge/

Any thoughts about making a port to PyTorch Mobile?

Probably. But PyTorch Mobile is still very experimental, so you're probably better off with TF Lite.

Incorrect results, mostly predicting 'box turtle' and 'cd player' for every image.

I haven't tried this myself but I see other people reporting similar issues (with CoreML and Vision).

Does the execution time includes image pre-processing?

In this app, the execution time also includes the time needed to schedule the Vision request (which in turn calls Core ML), the preprocessing, the neural network, the post-processing of...