Finetuning and Bulk Labelling IMAGES with Prodigy
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 how you might be able to improve the annotation experience by using Bulk Labelling. It's a technique we've used in an earlier video for text, and the goal of this video is to show you how you can use it for image classification as well.
[00:00] Introduction
[00:48] Theory behind Bulk Labelling
[02:39] Color Histograms
[05:44] Twitter Emoji Code
[08:20] Twitter Emoji Demo
[10:03] Color Histogram Weakness
[12:08] Convolution Models
[16:21] Pretrained Models and Code
[17:42] MobileNet
[18:41] Xception
[20:49] The Finetuning Trick
[23:58] Prodigy Mark Recipe
[26:05] Finetuning with Keras
[32:05] Finetuned Bulk Labelling
[34:42] Lessons Learned
PRODIGY
● Website & docs: https://prodi.gy
● Live demo: https://prodi.gy/demo
● Forum: https://support.prodi.gy
THIS TUTORIAL
● Bulk Labelling for Text: • Bulk Labelling and Prodigy
● Bulk Helper Library: https://github.com/koaning/bulk/
● Embedding Helper Library: https://github.com/koaning/embetter
● The code used in this tutorial: https://github.com/explosion/prodigy-...
FOLLOW US
● Explosion: / explosion_ai
● Subscribe to our newsletter: http://eepurl.com/hZyFbf
● We offer new services! spaCy Custom Solutions✨ https://explosion.ai/custom-solutions