Learn how to prepare training data and labels for AutoML Vision to create custom label output for photos.
- So the first step in working with AutoML for vision…is working with your data. Now, there are a couple of…points to note about the data that is designed,…at this point, to work with this API.…First, data, recommended by Google, are images "from the…real world" and they specifically say not x-rays,…hand drawings, or receipts, in their documentation.…They also recommend high-quality images, and they say…do not use grainy images, such as surveillance photos.…Maximum is 30 megabytes per image.…
As of this recording, the file types supported are JPEG,…PNG, GIF, BMP, WEBP, TIFF, and ICO.…Now, let's take a look at the photos that I have prepared.…You can see I have a folder called "naturePhotos",…and I've created a zip from it.…Underneath that folder, I have child folders with the…intended label names. So, here I have "Flowers", "food",…"sky", "sunset", "trees", and "vista".…And again I'll remind you that you need to have a minimum…of 10 labeled images per category, or per label, in order…to build a basic model. So, I have my zip file ready to go…
Author
Lynn Langit
Released
10/23/2018- Hosting options: Serverless, containers, and virtual machines
- Enabling the GCP ML AIs
- Preparing data with Cloud Dataflow and Dataprep
- Modeling predictions for images, video, text to speech, and cloud translation
- Machine learning with AutoML
- Advanced machine learning and deep learning
- Machine learning architectures
Skill Level Intermediate
Duration
Views
-
Introduction
-
What you should know1m 13s
-
About using cloud services1m 33s
-
1. Machine Learning on Google Cloud Platform
-
GCP AI servers vs. platforms5m 11s
-
Enable GCP ML APIs4m 25s
-
2. Machine Learning API Services
-
Overview of GCP ML APIs2m 38s
-
Predict via BigQuery ML6m 29s
-
-
3. Machine Learning with AutoML
-
Understand AutoML Vision4m 57s
-
4. Advanced Machine Learning
-
Why build custom ML models?6m 59s
-
Use Cloud ML Engine8m 2s
-
Scale custom ML models3m 47s
-
Understanding deep learning4m 33s
-
Work with TensorBoard4m 45s
-
GPUs and TPUs for TensorFlow5m 21s
-
-
5. Machine Learning Architectures
-
Chatbot with ML2m 42s
-
GCP ML service for IoT apps3m 10s
-
-
Conclusion
-
Next steps1m 30s
-
Take notes with your new membership!
Type in the entry box, then click Enter to save your note.
Press on any video thumbnail to jump immediately to the timecode shown.
Notes are saved with you account but can also be exported as plain text, MS Word, PDF, Google Doc, or Evernote.
Share this video
Embed this video
Video: Prepare data and labels for AutoML Vision