Thanks for all the information Nickolay. You've been very helpful. Vosk does seem to be a very good SRE, but the dictionary modifications appears to be too complex for my purpose. So I'm going to need to go back to using pocketsphinx just for the ease-of-use of its dictionary. In any case, thanks for your patience and time. You've been very helpful.
It can be difficult to get a simple answer to a question. I've got vosk installed and running on a computer. Vosk only. Not Kaldi. Although Vosk apparently comes with some form of a Kaldirecognitzer. My question is where do I find the vocabulary dictionary, and how do I modfied it? And is it easy to do? With pocketsphinx it is extremely easy to modify the dictionary or even created a new dictionary from scratch. And that was the main attraction to pocketsphinx. I need to be able to create a dictionary...
"There is C documentation here. https://cmusphinx.github.io/doc/pocketsphinx/files.html. There is no Python documentation, just the source code." This will be very helpful. Thank you! Thanks for the information on alsa config, and json too. Sorry to hear that there's no documentation for vosk. How's a person supposed to learn it? Just keep coming here and asking questions? I just want to iron out a few things so I can use an SRE for my Linquistic AI project. I like Pocket Sphinx because I understand...
Ok I finally got VOSK running, but as I suspected I have simlar questions with vosk: Here's my Python code: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37#!/usr/bin/python3 from vosk import Model, KaldiRecognizer import time import os os.system('clear') time.sleep(1) if not os.path.exists("model-en"): print ("Please download the model from https://github.com/alphacep/kaldi-android-demo/releases and unpack as 'model-en' in the current folder.")...
I did install libgfortran as a dependency Kaldi. I'll install libgfortran3 and see if that helps. Although I'm probably going to have as many questions about vosk as I have for pocketsphinx. The first thing I'll want to do is empty its dictionary and start my own dictionary from scratch. I already know how to do that for pocketsphinx. Currently I have pocketsphinx running. And it's decoding my speech close to 100% accuracy. In fact, most of the time it is 100% accurate. As I say, it must like my...
I'm not getting anywhere with vosk. I've installed vosk and Kaldi. I'm trying to run the following python text code for vosk 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29#!/usr/bin/python3 from vosk import Model, KaldiRecognizer import sys import os import wave if not os.path.exists("model-en"): print ("Please download the model from https://github.com/alphacep/kaldi-android-demo/releases and unpack as 'model-en' in the current folder.") exit (1) wf = wave.open(sys.argv[1],...
Ok, that seemed to work. james@james-desktop:~/vosk-kaldi$ pip3 install https://github.com/alphacep/vosk-api/releases/download/0.3.3/vosk-0.3.3-cp36-cp36m-linux_aarch64.whl Defaulting to user installation because normal site-packages is not writeable Collecting vosk==0.3.3 Downloading https://github.com/alphacep/vosk-api/releases/download/0.3.3/vosk-0.3.3-cp36-cp36m-linux_aarch64.whl (2.5 MB) |████████████████████████████████| 2.5 MB 4.9 kB/s Installing collected packages: vosk Successfully installed...
Ok, I have a problem right off the bat: I'm running Ubuntu 18.04 on a Jetson Nano arm64. From the GitHub page I tried the foillowing: james@james-desktop:~/vosk-kaldi$ pip3 --version pip 20.0.2 from /home/james/.local/lib/python3.6/site-packages/pip (python 3.6) (I have the correct verion of pip) james@james-desktop:~/vosk-kaldi$ pip3 install vosk Defaulting to user installation because normal site-packages is not writeable ERROR: Could not find a version that satisfies the requirement vosk (from...
"If it is perfect already, what are you asking about then?" I was asking for information on how to access and control pocketsphinx from Python. I wasn't complaining that it doesn't decode speech well. And I have a feeling that if I move over to vosk-kaldi I'll end up having the same questions. And I will have basically gained nothing. But hey, I'll give it a shot. The thing that is so disusting is that I have already spend about 2 weeks learning all about pocketsphinx. Now I'll need to start all...
Other people have pointed me to vosk as well. The problem is that I haven't been able to find much information about vosk. I don't even know what it is, or how to use it? Does vosk use pocketsphinx? If not, what exactly is vosk? And where do I find detailed information on it beyond that github page, especially in terms of tutorials.? I'm already having close to 100% accuracy with Pocket Sphinx. It's been decoding everything I throw at it with near perfection. Perhaps it likes the way I speek? Also,...
I have PocketSphinx installed on a small SBC running Ubuntu 18.04. I have also installed PocketSphinx-Python and I have been able to run some example python code I found including one that allows me to use microphone input. However, I would like to write my own Python programs to access pocketsphinx, but I don't know where to find the methods used by pocketsphinx-python or how to use them. For example there is a method called "decode" In the example program they are using decode.end_uut() I have...
I'm really liking eSpeak for my AI project. This appears to be exactly what I was hoping to find. What I like about it is that I can basically design how it speaks from the ground up which is exactly what I need for my project. Unfortunately I'm having difficulty finding good tutorials on exaclty how to work with eSpeak. What I'm doing right now is preparing a new dictionary that begins from scratch. So I want this dictionary to contain the absolute minimal number of words and phonemes. My Linguistic...
Just thought I'd add the following information for anyone who might be interested. I was trying to have eSpeak say, "my name is alysha". As addressed in the previous posts the name was being pronounced incorrectly and I corrected that by adding new phonemes for the name alysha. However, there was still a problem. In the sentence, for some reason the "is" that came right before alysha was causing an "m" sound to be produced which make alysha sound like "malysha" It too me a while to find where this...
Progress UPDATE again: First, for the sake of anyone else searching this forum I found the en_rules and en_list files in a directory named dictsource that was included in the originalk download zip file with espeak. In any case, I had some success. Here's what I did. MY PROBLEM. I wanted espeak to speak the name Alysha somewhat as Aleesha. It was currently saying it as Alyisha. I used espeakedit to see that the current phoneme translation of alysha was: 'aliS@ So then I typed in Alyeesha and it spoke...
UPDATE: I found the en_rules and en_list files. I can open them with a text editor. I tried to open them with espeakedit but all that opens are empty tabs? So I'm still at a loss as to how to modify these files and complie them back into an en_dict file. For right now all I want to do is add some words with special prounuciations into the en_list file and see how that works. I may be interested in working with the en_rules file too, but that would come later.
I'm using eSpeak in an AI project and I would like to be able to customize the way it prounces specific words. It's my understanding (which could be wrong) that this information is storied in the language dictionary file (i.e. en_dict for example). But how do I modify that file? I've installed eSpeak Edit and I've been trying to learn how to use it but it's difficult to find a good tutorial on it. It's my understand (which could be wrong) that the en_dict file is compled from two other files supposedly...
I'm just starting out with pocketsphinx for use on an AI project. I've downloaded and installed pocketphinx, sphinxbase, and sphinxtrain. The system appears to be working just fine from the Linux command line. But I don't know how to access it from Python3. Can anyone point me to any tutorials on how to use pocketsphinx from Python3? I currently have the system set up on a Raspberry Pi 4 using Ubuntu 18.04 OS. I'm using Code-OSS as the Python IDE. I can't seem to find any clear tutorials on how to...
Hello, my name's James a.k.a Robo Pi. I'm starting an A.I. project based on the semantics of words. I've decided to employ both TTS (Test to Speech) and SRE (Speech Recognition Engine) For the former I've chosen to use pyttsx3 as the TTS module for Python which is the language I'm using for the A.I. project. I've come to realize that pyttsx3 actually uses espeak as the underlying TTS driver. I should probably mention the system I"m using: I've chosen to create this project on a Raspberry Pi model...