search

maikherbig / AIDeveloper

GUI-based software for training, evaluating and applying deep neural nets for image classification
BSD 2-Clause "Simplified" License
109 stars 20 forks source link

Possibility for R-CNNs, SSDs and GANs #13

Open alfrem opened 4 years ago

alfrem commented 4 years ago

Would it mean massive changes to GUI to be able to implement that possibilities? Thanks!

maikherbig commented 4 years ago

For now, I want to focus on classification. This means, the input to the model is an image and a class (one single number) and the output is the prediction for the classes (as many numbers as there are classes). GANs are often used for image-to-image translation. For R-CNNs, the input is multiple images. Single shot detectors (SSDs) take and image and output a few values (bounding boxes)

None of these are impossible, but each would require quite some time to implement.