Closed marcverhagen closed 9 months ago
Did much of this in https://github.com/clamsproject/app-swt-detection/commit/eaa65220234120eecc1011979e8c03253494d4fb. But due to limitations imposed by mmif-python (see https://github.com/clamsproject/mmif-python/issues/252), the classification for timepoints looks like
"classification": [
"slate:4.1978837543865666e-05",
"chyron:0.9895544052124023",
"credit:0.0007274810341186821",
"NEG:0.009676150046288967"
]
Not sure what is worse, the above or
"labels": ["slate", "chyron", "credit", "NEG"],
"scores": [4.1978837543865666e-05, 0.9895544052124023, 0.0007274810341186821, 0.009676150046288967]
closing as duplicate to #41
Because
The MMIF output now is a list of TimeFrames:
After some discussion with @keighrim we are now thinking:
The targets would be TimePoints with scores for all labels, for example:
See also #41 .
Done when
No response
Additional context
No response