I would like to use the Google Videointelligence API (Speech detection, object detection) and process approx. 100 videos which are stored on Google Cloud Storage. Each video has an approximate length of 1min. I do not want to process them sequentially because this will take a while.
Is there any opportunity to send all the files to the Google Videointelligence API so that they are processed at once?
Regards, Andi
By specifying in the input link 'gs://BUCKETNAME/*' then it will read all files in the folder
Indeed, creating a bucket in Cloud Storage with all video files in it and referencing to its name in your Video Intelligence API’s file path will process all files at once.
Here is a quickstart on the Video Intelligence API where the gcloud command line references to a unique file:
In case you would be using a REST API, here is some information on the JSON representation for the annotate method for example, where the URI for the Cloud Storage bucket is referenced.
Additional information on the gcloud command’s input_path can be found here, using a detect-label command for example.
Finally, here is how to get information about your Cloud Storage buckets.
