Ad-Detective

Ad-Detective frame statistics and review

Interpret candidate evidence and use review controls to remove false positives before creating edits.

What the statistics describe

Candidate statistics summarise the analysed picture evidence used by scene and black-frame rules. They are diagnostic measurements, not a guarantee that a transition is an advertisement boundary.

Review in context

  1. Navigate to the candidate frame.
  2. Inspect several adjacent single-frame thumbnails and the waveform.
  3. Compare black/scene statistics with nearby ordinary dark or high-motion frames.
  4. Keep, remove or convert the candidate to a chapter mark/cut.

Improve the next scan

Patterns in false positives indicate which parameter to change: threshold for excessive scene marks, black/contrast/region settings for dark-content mistakes, or spacing/block rules for incorrect candidate grouping. Retest a short sample before rescanning a long source.