Get end-to-end BBQS workflow example running (labeling)

[rhingo]

Step 1 of 4 of running an end-to-end SLEAP workflow with actual BBQS data:

1. Labeling
2. Training
3. Inference
4. Visualization / Validation

[CodyCBakerPhD]

I think the top priority for this would be to schedule and perform user tests of the two interfaces on videos from the same dataset:

• https://app.sleap.ai/dev/
• https://codycbakerphd.github.io/pozu/

(I still need to transcode more videos and expose them to Pozu, but can get that done before any scheduled test)

Basically, document how smooth each workflow is, how many steps each takes, any shortcomings or bugs, and importantly how quick it is to go label frames once labeler 'gets into the zone' since that determines how many total work hours it would take a horde of labelers to generate a million training frames