Is it Food? An Introduction to Machine Learning

on YouTube ↗35m

" summary

A practical walkthrough shows an image classifier built with TensorFlow and Inception v3, retrained via transfer learning on folders of images (roughly 1–2k images per food class, ~600 for text), packaged as a Python Flask service that a Rails app calls. The retrain script runs slowly but produces a graph + labels and yields about 98% accuracy in production for their use case; code and pitfalls (Docker, compilers, async vs sync) are demonstrated.

AI-written summary · about

" chapters (17)

1. 0:00Intro
2. 0:12Who am I
3. 1:59Why this talk
4. 3:05Goals
5. 5:58Pattern Recognition
6. 8:54Tensorflow
7. 11:07Is it Food
8. 12:27Mistakes
9. 13:45Installation Setup
10. 15:06Transfer Learning
11. 17:48What are bottlenecks
12. 20:59Labeling an image
13. 23:58Building a web server
14. 24:20Using Flask
15. 25:20Using Rails
16. 27:02Additional Resources
17. 27:28Questions

" related talks

• 01h 00mA gentle introduction to Deep LearningLaurent Bugnion · NDC Porto 2022
• 01h 29mComputer Vision: An (Un?)Expected Journey, with Keras and TensorflowRodolfo Bonnin · PyData London 2018
• 55mTensorflow: Learning MachinesVikram Tiwari · Windy City DevFest 2018
• 38mReal-world Machine Learning with TensorFlow and Cloud AIKaz Sato · Heapcon 2019
• 40mMachine learning 101: Where to begin?Lutske de Leeuw · J-Fall 2023
• 50mBuilding Intelligent Mobile Apps using Machine LearningWei-Meng Lee · Mobile Era 2018
• 21mTransform your web apps with Machine LearningBhavesh Bhatt · MumbaiFOSS 2023
• 41mBeyond the Algorithm: What Makes Machine Learning Work?Ellen Friedman · Berlin Buzzwords 2018