IMAGE CAPTIONING ANNOTATION with Prodigy & PyTorch: custom scriptable machine learning annotation
Prodigy is a modern annotation tool for collecting training data for machine learning models, developed by the makers of spaCy. In this video, we'll show you how you can use Prodigy to script fully custom annotation workflows in Python, how to plug in your own machine learning models and how to mix and match different interfaces for your specific use case. We'll create a dataset of image captions, use an image captioning model implemented in PyTorch to suggest captions and perform error anylsis to find out what the model is getting right, and where it needs improvement.
STEP BY STEP
01:59 – Create a recipe for manual image captioning
13:26 – Plug in an image captioning model implemented in PyTorch and correct its output
19:41 – Add callback to count changed captions and output results
26:33 - Add workflow for error analysis and reviewing annotations
PRODIGY
● Website & docs: https://prodi.gy
● Live demo: https://prodi.gy/demo
● Custom recipes docs: https://prodi.gy/docs/custom-recipes
● Forum: https://support.prodi.gy
● Recipe scripts: https://github.com/explosion/prodigy-...
THIS TUTORIAL
● Code: https://github.com/explosion/prodigy-...
● PyTorch tutorial & model: https://github.com/yunjey/pytorch-tut...
● Images dataset: https://www.kaggle.com/alessiocorrado...
FOLLOW US
● Ines Montani: / _inesmontani
● Explosion: / explosion_ai
CREDITS
● Thumbnail cat image #1: https://unsplash.com/photos/E9kVmtiqqGE
● Thumbnail cat image #2: https://unsplash.com/photos/_Mty3g9XWr0