Ready
Awaiting processed results
The job is queued or running. Progress updates will appear here as each image or video is processed.
CONCEPTFlow MEDIA DESCRIBER
MEDIA DESCRIBER: CONCEPTFlow Native App
Upload one or more images and/or videos. Each input is resized to 896p, processed one at a time, and resulting OCR, VISION and TRANSCRIBER outputs are rendered as results flow.
Media Describer build request. Prompt ready.
- Awaiting progress updates.
Pipeline summary
- Queue each uploaded file, then process it sequentially in the background.
- Resize every image or video frame to a maximum edge of 896 pixels while preserving aspect ratio.
- For videos, extract frames at 2 fps and skip frames that are at least 97% similar to the previous kept frame.
- Run OCR and VISION on every kept image or frame, then transcribe audio segments centered on each kept video frame when speech is available.
- Render the final result in an accessible result region with focus support for screen readers.