Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Investigate image auto-tagging #134

Open
rlnorthcutt opened this issue Jul 24, 2019 · 5 comments
Open

Investigate image auto-tagging #134

rlnorthcutt opened this issue Jul 24, 2019 · 5 comments

Comments

@rlnorthcutt
Copy link
Contributor

We want to have media auto tagging (via ML API) as an option. It won't be enabled by default because it will require an API key. The service needs to provide a free tier with reasonable limits, and ideally has good pricing for those who want to keep using it.

When enabled and configured, the module will automatically add image tagging for the media_tag taxonomy on media.

So - the purpose of this issue to research and identify the right service to settle on, and get a rough idea of the architecture. We can create a followup issue with the architecture and dev plan.

Overview of APIs
https://medium.com/activewizards-machine-learning-company/comparison-of-top-6-cloud-apis-for-computer-vision-ebf2d299be73

Example D8 module for Google Cloud:
https://github.com/kalamuna/google_cloud_vision

@saltednut
Copy link
Contributor

Does this make the boat go faster? Seems like a nice to have but would never come up in a demo?

@rlnorthcutt
Copy link
Contributor Author

We have been asked about this a few times (and ML in general). However, I do think it makes the boat/car faster because it fills out form fields for you.

Ideally, SAs add media taxonomy when creating media. This then allows them to filter/search by that taxonomy in the media browser. The story is media discovery and organization.

If we can automate this for SAs, and once we finish #132, then all an SA really has to do is upload an image, and most of the media fields are filled out by default without extra effort.

That would make a noticeable improvement in both speed as well as completeness.

@saltednut
Copy link
Contributor

once we finish #132

Is this actually blocking 132 or are we just making reference to it?

@rlnorthcutt
Copy link
Contributor Author

Just making reference. The dream machine is that I upload an image, and then all the metadata (name, alt text, tags, etc) is done for me in a useful way.

@saltednut
Copy link
Contributor

Thanks just making sure. I am riding some kind of negative wave right now so I figured it a good time to comb over the existing issues and nitpick out the details. Don't worry, I'll be alright. 😱

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

2 participants