Question about using bounding boxes in training data for Cloud AutoML Video Intelligence Classification

[Jun Heider]

Hello,

So I've taken a look at the documentation: https://cloud.google.com/video-intelligence/automl/docs/ and this google group. It's apparent how to create training data with annotations labels for the segments of the videos "object X for start time to end time" but I'm not sure how to create training data CSVs with bounding boxes that Cloud AutoML Video Intelligence Classification will parse and use during training? I would assume the bounding box would be crucial to help train the model on what region the annotation represents.

Is there a tutorial or examples somewhere on what the bounding box input data for the training dataset should look like?

If I missed something in the documentation, I'll be the first to admit I did once it's pointed out, but I didn't see any sample dataset examples that I recall that illustrate bounding box.

Thanks in advance!

[Alex Chen]

Hi Jun,

I assume you are actually looking for object tracking document. The doc you followed is for classification.

Here is the object tracking doc link for how to prepare csv files including format of bounding boxes:

https://cloud.google.com/video-intelligence/automl/object-tracking/docs/prepare#csv

The quickstart has a sample data set for object tracking where you can find the format of csv files. Please take a closer look of all related csv files.

https://cloud.google.com/video-intelligence/automl/object-tracking/docs/quickstart-console

Feel free to reach out if you have any questions.

thanks,

-Alex

--
© 2018 Google Inc. 1600 Amphitheatre Parkway, Mountain View, CA 94043
 
Email preferences: You received this email because you signed up for the Google Cloud Video Intelligence API Discussion Google Group (cloud-videoi...@googlegroups.com) to participate in discussions with other members of the Google Cloud Video Intelligence API community and the Google Cloud Video Intelligence API Team.
---
You received this message because you are subscribed to the Google Groups "cloud-videointel-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cloud-videointel-d...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cloud-videointel-discuss/e7d921ff-9c6a-4d10-86a7-623d2a3c771c%40googlegroups.com.

[Jun Heider]

Alex,

Much appreciated, I will try out the object tracking version of auto ml. I will reach out if any questions, but thanks for pointing me in the right direction!

To unsubscribe from this group and stop receiving emails from it, send an email to cloud-videointel-discuss+unsub...@googlegroups.com.