Open peetucket opened 1 month ago
I think we are currently processing a video with no content, but it would be nice to test something that actually has VTT results that can be checked?
Agreed. Suggestions for video content that is real enough to have some text to transcribe but not too long?
Verify the speechToTextWF works as expected for audio and video via accessioning from pre-assembly through to purl, analogus to the ocr specs for images and document.
Useful to have early on so we can easily push objects through the workflow and find where it's still broken