Alexandre Carlier

Hey, great question! I'll write a simple notebook this evening or tomorrow explaining the process step by step. 1. You're correct, you need to have individual glyphs in SVG format....

Will close when I've written the custom dataset creation notebook 😉

Hey, thanks! Great suggestion! This was definitely on my todo-list since, even for people with programming experience, running the experiments requires a CUDA GPU. I don't have access to my...

Hello @tsaxena! Yes unfortunately this is due to the fact that the pretrained model was trained with a maximum of 8 paths per SVG. Since an index embedding is used...

Yes, the dataset is about 100k icons, but because of time constraints the pretrained model was only trained on a filtered subset. And I don't have access to the GPU...

Hey Alex, Yes for sure, we even had this kind of application in mind while developing DeepSVG. Instead of training for input reconstruction, just replace the output target with the...

Great! - While 1. sounds potentially feasible, I see the problem that you won't have enough training data and the model will just overfit without capacity to generalize. - 2....

Hello @tsaxena, I recently added a Dataloader which can directly handle SVG icons as input. You can first pre-process the entire dataset using the https://github.com/alexandre01/deepsvg/blob/master/dataset/preprocess.py script or load https://github.com/alexandre01/deepsvg/blob/master/deepsvg/svg_dataset.py with...

Let me just give some clarification regarding filling. Let's take the example of a filled "donut" shape: a filled circle with a "hole" inside. This is done in SVG by...