Can't recognize my words

DeepSpeech
#1

input wav file:
s1.wav.zip

model_file = ‘deepspeech-0.9.3-models.pbmm’
scorer_file = ‘deepspeech-0.9.3-models.scorer’

output:

{
  "transcripts": [
    {
      "confidence": -17.701541900634766,
      "words": [
        {
          "word": "i",
          "start_time": 3.76,
          "duration": 0.0
        }
      ]
    }
  ]
}

something wrong with me?

#2

The audio file sounds Chinese. The model file and scorer and output look English.

#3

“I have a book”

Maybe I’m not reading it clearly. :sweat_smile:

#4

Oh, sorry, I think I must have seen the Chinese title and got confused. Yes , I hear it now, but I also see that the file is stereo. You want to try and give it mono input.

#5

Thanks for your reply.
I try “ffmpeg -i s1.wav -ac 1 s2.wav” get mono file.
It’s work!

2 Likes