Directly feeding sampled desktop-recording frames to a vision-language model extracts click/select/scroll/drag/type action sequences more reliably than explicitly computing frame differences first.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Sharingan: Extract User Action Sequence from Desktop Recordings
Directly feeding sampled desktop-recording frames to a vision-language model extracts click/select/scroll/drag/type action sequences more reliably than explicitly computing frame differences first.