Text extraction

Job request to extract text from some kind of input.

Input

Jobs MIGHT specify a mime type as the marker in the i tag.

Params

range

Specifies a range on which to focus the transcription in seconds (for audio inputs)

[ "param", "range", "<start>", "<end>" ]

alignment

Specifies alignment of labels (for audio inputs). For example, if the output format supports timestamps, these will occur according to the alignment. If parameter is not set, the DVM decides the default alignment. possible inputs raw, segment, word

[ "param", "alignment", "raw" ]

Output

Including but not limited to:

  • text/vtt
  • text/plain
  • text/markdown

Example

{
    "content": "",
    "kind": 5000,
    "tags": [
        [ "i", "https://chtbl.com/track/E2818G/http://feeds.soundcloud.com/stream/1602940803-bitcoin-and-data-vendetta-ep782.mp3", "url" ],
        [ "output", "text/plain" ],
        [ "param", "range", "300", "360" ]
        [ "param", "alignment", "raw"]
    ]
}