直接从 SFTP 服务器将音频文件加载到 Python 的语音识别模块（使用 Paramiko SFTPClient）

Question

I want to preface this with the fact that I am still very much a noob at Paramiko, so this might all be completely impossible.

I want to open a .wav file on a server from my computer in order to do some speech recognition on it. To do this, I create a Transport with Paramiko and use it to open the audio file on the server. Then I set this file as my source and use SpeechRecognition to print whatever's being said in the audio file. However, when I open the file located in remotefilepath , it is no longer recognised as an audio file, since I get the error message AssertionError: Source must be an audio source . Printing type(file) I get <class 'paramiko.sftp_file.SFTPFile'> .

I want to just open the file on the server and do the speech recognition on it without having to first save the file onto my own computer. Is there any way for me to do this? Any help/advice is very much appreciated

import paramiko
import speech_recognition as sr

remotefilepath = /path/to/file.wav
server_ip = 12.34.567.8
server_port = 22
transport = paramiko.Transport((server_ip, server_port))
transport.connect(username="foo", password="bar")
print "Connected to transport"
sftp = transport.open_sftp_client()
file = sftp.open(remotefilepath)

#print type(file)

r = sr.Recognizer()

with file as source:
    r.adjust_for_ambient_noise(source)
    audio = r.record(source)
    try:
        text = r.recognize_google(audio)
        print "You said: {}".format(text)
    except:
        print "Sorry, I could not understand."

file.close()
sftp.close()
transport.close()
print "Closed transport. Ending program" 

Python 2.7 on Windows 10.

Answer 1

adjust_for_ambient_noise takes an implementation of AudioSource as an argument. You are passing file-like object instead.

I believe you can use AudioFile implementation of AudioSource instead. AudioFile can be created with file-like object:

with sftp.open(remotefilepath) as file:
    with AudioFile(file) as source:
        r.adjust_for_ambient_noise(source)

Answer 2

尝试将文件从源复制到本地并开始处理它。

sftp.get(filepath, localpath)