Recherche avancée

Sur d’autres sites (5507)

• Google Speech API doesn't give correct result when audio is sent in file

  4 août 2012, par Cupidvogel

  I chanced upon the article at Google Speech API which suggested a mechanism for extracting text from audio file through Perl. Now I have recorded a audio file, which you will find at http://vocaroo.com/i/s0lPN5d3YQJj. It is a simple piece of audio, reading I love you. When I go to the Google speech API in Chrome, and speak those words, I get the right result. When I try the code at the above mentioned link with the audio file I pointed out, it returns strange results, like logan. How can I make it more accurate ? This is just a sample audio, what I am generally doing is extracting the audio from a video file through FFMpeg using something like ffmpeg -i input.avi -vn -ar 44100 -ac 2 -ab 192 -f mp3 output.mp3, followed by ffmpeg -i input.mp3 output.flac.

• Improving Google Cloud Speech-to-Text accuracy

  6 juillet 2020, par lr_optim

  I'm working on a project where I need to perform these steps :

  


  

  1. Record a voice call (.webm -file)

  


  2. Split the webm -file into chunks with ffmpeg and convert the file into wav

  


  3. Transcribe the chunks using SpeechRecognition -library and Google Cloud API

  


  


  I've faced problems with the transcription accuracy and wondering if there is something I could do to improve it. At the time I'm splitting the original file into 30s chunks. I thought there might be one problem, that I might be missing words because of splitting so I've tried also with longer chunks under 60s but didn't notice any improve in accuracy.
Reading trough the speechRecognition docs I decided to set r.energy_threshold = 4000, I also tried to set the energy_treshold dynamically like this :

  


  with sr.AudioFile(name) as source:
    r.dynamic_energy_threshold = True
    r.adjust_for_ambient_noise(source, duration = 1)
    audio = r.record(source)


  


  I've also tested en-US and en-GB to see if there's some difference but there isn't as much as I'd want. The program is supposed to work with english language spoken by nordic people. If someone has experience about choosing a right language model for people speaking with accent, please let me know.

  


  This is the ffmpeg command is use to split the webm file into chunks : command = ['ffmpeg', '-i', filename, '-f', 'segment', '-segment_time', '30', parts_dir + outputname + '%09d.wav']

  


  Is there somethig I could do better ? I'm wondering if the quality is not good enough an Google is having hard time because of that ?

  


  The main problem is I'm getting bad results (lots of wrong words) from Google and wondering if there is something I could do about it.

  


• ffmpeg won't execute properly in google app engine standard nodejs

  3 septembre 2019, par tommyc38

  I have tried for three full days to get GAE (standard - nodejs) to run a simple video transcoder from MOV to MP4 using ffmpeg. I have tried using ffluent-ffmpeg, kicking off a child process (e.g. spawn), and nothing works. As soon as it hits the call to the executable it always errors. I have confirmed ffmpeg is installed and even tried using ffmpeg-static. Moreover, I have it working on my local machine with no problems (using all of the aforementioned ways).

  I have also tried logging the errors and nothing is really all that helpful. I can see its working through any installed package including ffmpeg (system package).

  Below is the pseudo code...step three is where the problem occurs.

  1. Send file name to GAE endpoint
  2. Download the file from google cloud storage to a temp file
  3. Transcode using ffmpeg
  4. Upload temp file to google cloud storage
  5. Remove old google cloud storage file
  6. Remove temp file

  The file I am using to test is 6MB...a 5 second video I took on my iPhone. Thank you in advance.

  UPDATE : I successfully deployed the exact same code to Node Flex environment and everything works great. I wasn’t able to get any errors in the standard environment that directed me where to look but my guess is it has something to do with how it stores the file I pipe into FFMPEG on GAE Node Standard. The docs say its a virtual file system that uses RAM. I’d love to hear if anybody managed to get it working in the standard environment.