从头开始简单的语音识别 [英] Simple speech recognition from scratch

问题描述

我发现与我的问题最相似的问题是这个(简单的语音识别方法)，但由于已经过去3年了，所以我问的答案还不够.

The most alike question I found related to my question is this (simple speech recognition methods) but since had passed 3 years and the answers are not enough I will ask.

我想从头开始计算一个简单的语音识别系统，我只需要识别五个单词即可.据我所知，此应用程序最常用的音频功能是MFCC和用于分类的HMM.

I want to compute, from scratch, a simple speech recognition system, I only need to recognize five words. As much as I know, the more used audio features for this application are the MFCC, and HMM for classification.

我能够从音频中提取MFCC，但对于如何使用这些功能通过HMM生成模型然后进行分类，我仍然存有疑问.

I'm able to extract the MFCC from audio but I still have some doubts about how to use the features for generating a model with HMM and then perform classification.

据我了解，我必须执行矢量量化.首先，我需要有一堆MFCC向量，然后应用聚类算法来获取质心.然后，使用质心执行矢量量化，这意味着我必须比较每个MFCC矢量，并以最相似的质心名称对其进行标记.

As I understand, I have to perform vector quantization. First I need to have a bunch of MFCC vectors, then apply a clustering algorithm to get centroids. Then, use the centroids to perform vector quantization, this means that I have to compare every MFCC vector and label it with the name of the centroid most alike.

然后，质心是HMM中的可观察符号".我必须将单词介绍给训练算法，并为每个单词创建一个HMM模型.然后，在进行音频查询时，我将所有模型进行比较，然后说这是概率最高的单词.

Then, the centroids are the 'observable symbols' in the HMM. I have to introduce words to the training algorithm and create a HMM model for each word. Then, given an audio query I compare with all models and I say is the word with the highest probability.

首先，此过程是否正确?然后，如何处理大小不同的单词.我的意思是，如果我训练了500ms和300ms的单词，我会引入多少个可观察符号以与所有模型进行比较?

First of all, is this procedure correct? Then, how do I deal with different sized words. I mean, If I have trained words of 500ms and 300ms, how many observable symbols do I introduce to compare with all the models?

注意:我不想使用狮身人面像，android API，microsoft API或其他库.

Note: I don't want to use sphinx, android API, microsoft API or other library.

注意2:如果您共享更多最新信息以寻求更好的技术，我将不胜感激.

Note2: I would appreciate if you share more recent information for better techniques.

推荐答案

首先，此过程正确吗?

First of all, is this procedure correct?

矢量量化部分还可以，但是最近很少使用.您描述了没人使用的所谓离散HMM.如果您希望使用GMM作为排放的概率分布的连续HMM，则不需要矢量量化.

The vector quantization part is ok, but it's rarely used these days. You describe so-called discrete HMMs which nobody uses for speech. If you want continuous HMMs with GMM as probability distribution for emissions you don't need vector quantization.

然后，您专注于不太重要的步骤，例如MFCC提取，但是跳过了最重要的部分，例如使用Baum-Welch进行HMM训练和使用Viterbi进行HMM解码，这比通过矢量量化对状态的初始估计要复杂得多.

Then, you focused on less important steps like MFCC extraction but skipped most important parts like HMM training with Baum-Welch and HMM decoding with Viterbi which are way more complex part of the training than initial estimation of the states with vector quantization.

然后，我该如何处理大小不同的单词.我的意思是，如果我训练了500ms和300ms的单词，我会引入多少个可观察符号以与所有模型进行比较?

Then, how do I deal with different sized words. I mean, If I have trained words of 500ms and 300ms, how many observable symbols do I introduce to compare with all the models?

如果解码语音，通常会选择与人类感知的部分音素相对应的符号.传统上每个音素采用3个符号.例如，单词一个"应具有3个音素的9个状态，而单词七个"应具有5个音素的15个状态.事实证明，这种做法是有效的.当然，您可以略微改变此估算值.

If you decode speech you usually select the symbols which correspond to parts phonemes perceived by the human. Its traditional to take 3 symbols per phoneme. For example word "one" should have 9 states for 3 phonemes and word "seven" should have 15 states for 5 phonemes. This practice is proven to be effective. Of course you can vary this estimation slightly.

这篇关于从头开始简单的语音识别的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！