Lukas Muttenthaler

source-specific extractor classes that inherit from a base extractor class

1

- now, one can only use `DEFAULT` weights for `torchvision` which is limited behavior (our goal is flexible behavior) - we need to make feature extraction more flexible for a...

README improvement

2

Improve readability of README file. Make text more concise and easier understandable?

Custom models

1

Allow a user to add their own (pretrained) models without the necessity to upload them to `thingsvision` / make a PR - pass a model object to the extractor class...

69 extractor class for each source with inheritance from a base extractor class

1

- modularized extraction - base extractor class from which all source-specific extractor classes are supposed to inherit - mixin classes for both `PyTorch` and `TensorFlow` - source-specific extractor classes that...

Move everything model-/source-specific to custom model file or source-specific extractor

3

Currently, we have some if-statements that address model-specific exceptions. Since these are exceptions rather than something general, we want to specify them in the custom model file or move them...

Add source-specific config classes (e.g., `TimmConfig()`, `VisslConfig()`) similarly to how `huggingface` does it

- add models from the [`vissl`](https://vissl.ai/) library (in particular SimCLR) - we might want to do this in accordance with the fix for issue #69, since [`vissl`](https://vissl.ai/) will have its...

Library for feature extraction for Keras models

2

I just wanted to let you know that we are working on a library that facilitates feature extraction (i.e., extraction of image representations) for Keras models. :) Here's the link...

There are a few changes I would like to see, myself, but this should ultimately be Lukas's decision. I think we should ask for an `out-folder` and automatically name it...