The following technologies are relying on the Baidu big guy to provide interface operation! The calling interface language is Python
One, download
Pip Install Baidu-aip
Second, speech synthesis
According to the Baidu language synthesis official website, calls can:
#!/usr/bin/env python#-*-Coding:utf8-*- fromAipImportAipspeech"""your APPID AK SK"" " # After registering your Baidu account, create an app to get this dataapp_id="14446029" Api_key="ferct9kmokugdpbzqdqhqb7v"Secret_key='EXTB37XV1LUBXB8D5FZZNLLJLHNC7VMX'Client=Aipspeech (app_id, Api_key, secret_key) text="What's your name?"Res= Client.synthesis (text, options={ "Vol.": 8, # "Pit": 6, "SPD": 5, "per": 4})if notisinstance (res, dict): With open ('Audio.mp3','WB') as F:f.write (res)
Configuration parameters See: Http://ai.baidu.com/docs/#/TTS-Online-Python-SDK/top
This way we generate the ' audio '. mp3 file is the recognized language!
Third, language recognition
At present, Baidu supports the language format: the original PCM recording parameters must conform to the 8k/16k sampling rate, 16bit bit depth, mono, supported by the format: PCM (not compressed), WAV (uncompressed, PCM code), AMR (compressed format) (Baidu said);
In fact, Baidu can accurately identify the voice format only pcm!!!
Then we need to convert our voice format to PCM format! Use this tool to complete the audio in any format!
FFmpeg System Tools: https://pan.baidu.com/s/1jonSAa_TG2XuaJEy3iTmHg Password: W6HK
After the download, unzip, and then configure the environment variables, you can use it! Use the command to see the code below!
#!/usr/bin/env python#-*-coding:utf-8-*-ImportOS fromAipImportAipspeech"""your APPID AK SK"""app_id="14446029"Api_key="ferct9kmokugdpbzqdqhqb7v"Secret_key='EXTB37XV1LUBXB8D5FZZNLLJLHNC7VMX'Client=Aipspeech (app_id, Api_key, Secret_key)defget_file_content (file_path): Cmd_str= f"Ffmpeg-y-i {file_path}-acodec pcm_s16le-f s16le-ac 1-ar 16000 {file_path}.pcm"# for converting our voice formatsOs.system (CMD_STR) with open (f"{FILE_PATH}.PCM",'RB') as FP:returnFp.read () Res= Client.asr (Speech=get_file_content ("2.m4a"), options={ "Dev_pid": 1536,})Print(RES)Print(Res.get ("result") [0])
Parameter configuration see: Http://ai.baidu.com/docs/#/ASR-Online-Python-SDK/top
Four, natural language processing (NLP)
See: Http://ai.baidu.com/docs/#/NLP-Python-SDK/top
fromAipImportAIPNLP"""your APPID AK SK"""app_id="14446029"Api_key="ferct9kmokugdpbzqdqhqb7v"Secret_key='EXTB37XV1LUBXB8D5FZZNLLJLHNC7VMX'nip_client=AIPNLP (app_id, Api_key, Secret_key)defnip_handle (text):if nip_client.simnet("What's your name?", text). Get ("score") >= 0.72:#used to return two text similarity, greater than 72% can be considered, 2 meaning basically the same! return "My name is your little cutie." returnGoto_tuling (Text,"Xipangzi")
The Voice of artificial intelligence