@MODCAF footage to OpenAI Narrator via Python API

Подробнее о видео

iPhone footage from the @MODCAF TESTCHAMBER project into gpt-4o (small) model scene analysis which was in turn asked to generate a voiceover narrative in the style of David Attenborough and subsequently turned to speech using the Fable TTS model.
=============================================================
83 lines of python code via the OpenAI API
40.7 seconds from analysis to voiceover
(not including making the orginal video source)
=============================================================

OpenAI Generated Description:
----------------------------------------------------------
"Step into the unknown with this gripping video that delves into the depths of darkness and exploration. In a world suspended in shadows, a solitary figure armed with only a flashlight navigates an eerie, vast chamber."

OpenAI TTS Onyx model generated voiceover:
------------------------------------------------------------------------------
"In the depths of this colossal, dark chamber, a solitary figure ventures forth, illuminated only by the soft glow of a single light.

The door, ajar, offers a glimpse of sanctuary and safety, yet within this formidable expanse, darkness reigns supreme, devouring any semblance of illumination.

As he moves, the light dances on the cold, distant walls, revealing the vast and empty void that stretches endlessly around him.

This is a world few dare to enter, where shadows hold secrets, and silence is absolute. But even here, in this abyss, there is life and discovery, waiting to be unveiled by the curious and the brave."