Oguzhan Gencoglu

ONNX export

5

Do you support ONNX export and is it tested? Any examples of it would be appreciated.

Is this project abandoned?

2

Bunch of open PRs, no replies to any issues or discussion questions in the Discussion section, no releases...

How does this work with ONNX export and quantization?

6

Does quantized models here become quantized models in ONNX after conversion? Can you even convert/export them to ONNX? How about other way around? Can you export a sparse model to...

Feature: Sparse Matrix Multiplications

3

### Describe what you are looking for In numpy/scipy, `A @ B` is much faster if `A` is sparse and turned into scipy sparse matrix. If both are sparse or...

Support for python 3.11, 3.12

2

Any plans for support for newer python versions? Pypi does not seem to have any download files for 3.11 or 3.12

Speed benchmarks vs scipy

2

And speed benchmarks vs scipy / opencv / scikit-image for convex hull?

Does 2D fill has the concept of scipy's 'structure' ?

1

In `scipy` `binary_fill_holes`, one can control the definition of "hole" with e.g. `structure=ndimage.generate_binary_structure(2, 2)`. This specific example will produce  vs fill_voids fill:  The reason is the diagonal pixel...

Implementation in albumentations API

1

Thanks for the interesting work! It would be quite beneficial if these methods can be provided as a ready to use class that follows albumentation API which is probably the...

Any plans for support for newer python versions? Pypi does not seem to have any download files for 3.11 or 3.12

Thanks for your library. ONNX supports scikitlearn, lightgbm and xgboost models through `sklearn-onnx` converter. Inference speed benchmarks would be interesting and valuable. Any plans for that?