| Excerpt |
|---|
Verity processes videos by applying computer vision analysis to sampled image frames, plus text analysis of the video’s transcribed audio track. |
...
standalone audio tracks (or audio tracks extracted from video) by transcribing the audio to text. The speech-to-text output is enriched with any available |
...
metadata (such as title |
...
and description) |
...
then sent to |
...
Verity’s Natural Language Processing (NLP) machine learning models for classification. |
...