Implement split‑by‑sentence ( -sbs ) processing for bitmap subtitles
Implement split‑by‑sentence ( -sbs ) processing for bitmap subtitles: a task in LegoFlow-SWE (Harbor dataset). When extracting subtitles from bitmap sources (e.g., DVB bitmap captions), each output subtitle entry contains the text displayed at that moment. Because bitmap captions often accumulate…
The task
When extracting subtitles from bitmap sources (e.g., DVB bitmap captions), each output subtitle entry contains the text displayed at that moment. Because bitmap captions often accumulate text word‑by‑word across successive display intervals, the output contains **significant redundancy** – identical text is repeated…
Part of Lego-X/LegoFlow-SWE.