Last call to make your voice heard! Our 2022 Developer Survey closes in less than a week. Take survey.

Questions tagged [speech-to-text]

The tag has no usage guidance.

Filter by
Sorted by
Tagged with
0 votes
0 answers
11 views

Adding words to vocabulary on pre-trained ASR model

I have a pre-trained ASR model but want to add some missing words to the vocabulary. Can I do this or will it invalidate the entire training? Lets say I use the pretrained model: wav2vec2-base-960h ...
user avatar
  • 1
0 votes
0 answers
15 views

Using synthetic dataset for training NVIDIA NeMo Matchbox

Does anyone has success in training small command recognition models on synthetic dataset? The full details is the following: I need a small model to run a command recognition (about 30 commands) on ...
user avatar
0 votes
0 answers
11 views

Difference between speech recognition and automatic speech recognition

I was wondering, is there a difference between Speech Recognition and Automatic Speech Recogntion? I have seen both terms used in various papers, and I am not sure whether they are simply used ...
user avatar
  • 1
0 votes
2 answers
27 views

NameError: name 'librosa' is not defined [closed]

i'm working on Arabic Speech Recognition using Wav2Vec XLSR model. While fine-tuning the model it gives the error shown in the picture below. i can't understand what's the problem with librosa it's ...
user avatar
0 votes
0 answers
20 views

How can I build my voice speech-to-text model?

I found an instruction to build such kind of custom model on Azure. Prepare data for Custom Speech However, I would like to either fine-tune or transfer learning on Google Colaboratory or docker. In ...
user avatar
  • 119
0 votes
1 answer
20 views

Evaluate Text-to-speech without Human Involved?

I've explored text-to-speech evaluation matrices and they seem to used Mean Opinion Score (MOS) to evaluate a particular model. This matrice required humans to help to judge the model based on a scale ...
user avatar
0 votes
0 answers
58 views

NeMo Conformer-CTC Predicts Same Word Repeatedly When Fine-Tuning

I'm using the NeMo Conformer-CTC small on the LibriSpeech dataset (the clean subset, around 29K inputs, using 90% for training and 10% for testing). I use Pytorch Lightning. When I try to train, the ...
user avatar
0 votes
0 answers
33 views

How Pretraining part actually work in Wav2vec models? Which data is qualify to be the adequat for fine-tuning part the model of speech2text

Pretraining and fine-tuning the algorithm of wav2vec2.0, the new one using in FAcebookAI to do speech to text for low-resource language. I didn't actually get how the model does the pretraining part ...
user avatar
0 votes
0 answers
21 views

How to train my model on unpaired speech and text for my speech recognition model?

When we have unpaired data, means we have a dataset of audios and a dataset of texts, BUT they are not associated, and as we know, to build a speech recognition model we do have to combine in our ...
user avatar
0 votes
0 answers
32 views

Hidden Markov models in Speech Recognition

My first question here. So I am trying to build a sign language translator(from signs to text) and noticed that the problem itself is quite similar to speech recognition, so I started to research ...
user avatar
0 votes
0 answers
14 views

What are the hidden and observed states when building an acoustic model?

I have been trying to learn how to build ASRs and have been researching for awhile now, but I can't seem to get a straight answer. From what I understand, an ASR requires an Acoustic Model. That ...
user avatar
3 votes
1 answer
36 views

How to double audio dataset?

I am trying to develop a mispronunciation detection model for English speech. I use TIMIT dataset, this is phoneme labeled audio dataset. A phoneme is any of the perceptually distinct units of sound. ...
user avatar
1 vote
0 answers
18 views

How do I initialize a Hidden Markov Model when using MFCC features for speech recognition?

I have a personal dataset of 10000 audio files, each consisting a single spoken sentence. These files each have the transcribed text labels with them that I can use for supervised HMM training. Now ...
user avatar
  • 11
1 vote
1 answer
215 views

How does Wav2Vec 2.0 feed output from Convolutional Feature Encoder as input to the Transformer Context Network

I was reading the Wav2Vec 2.0 paper and trying to understand the model architecture, but I have trouble understanding how audio raw inputs of variable lengths can be fed through the model, especially ...
user avatar
0 votes
0 answers
31 views

How to build a speech corpus for continuous speech?

I am developing a speech corpus from audiobooks. I was able to collect recordings with their transcriptions but without timing information. Therefore I am looking for recent techniques to ...
user avatar
2 votes
2 answers
95 views

How to evaluate the quality of speech-to-text data without access to the true labels?

I am dealing with a data set of transcribed call center data, where customers are being recorded when interacting with the agent. This is then automatically transcribed by an external transcription ...
user avatar
2 votes
0 answers
76 views

How to do phoneme segmentation using dynamic time warping?

Background Information: Dynamic Time Warping (DTW): In time series analysis, dynamic time warping (DTW) is one of the algorithms for measuring similarity between two temporal sequences, which may ...
user avatar
2 votes
1 answer
43 views

How is an ASR's output compared to ground truth for validation?

I am curious how it is done as I am interested in doing something similar. I have some manually transcribed data that contains tags for multiple speakers. I want to compare how well the out of the box ...
user avatar
  • 319
0 votes
1 answer
35 views

About building database for persian speech to text project [closed]

In connection with the persian speech to text project on github, I am trying to collect the database from the audio and text available on the ganjoor site and Youtube Persian videos with Persian ...
user avatar
0 votes
1 answer
2k views

AttributeError: 'Functional' object has no attribute 'uses_learning_phase'

I am trying to predict on my test set and I get the following error AttributeError: 'Functional' object has no attribute 'uses_learning_phase' The model is created ...
user avatar
0 votes
0 answers
218 views

How to change voice features in python without affecting speech/language features?

I am trying to build a CNN model which should be able to identify the language being spoken in an audio file. I have extracted the MFCC matrix (for 13 coefficients) for each audio file and trained it....
user avatar
1 vote
1 answer
208 views

How to prepare Audio-text data for speech recognition

I have gathered some raw audio from all the conferences, meetings, lectures & casual conversation that I was part of. The machine transcription did not offer good results (from Azure, AWS etc.) I ...
user avatar
  • 11
0 votes
1 answer
633 views

ValueError: Error when checking input: expected the_input to have 3 dimensions, but got array with shape (14174, 1)

hope you're all doing good ! I am working on Automatic Speech Recognition with Python with the LibriSpeech Dataset. After preprocessing the audios data and applying an "MFCC featurizing" I ...
user avatar
1 vote
0 answers
22 views

How to make sense of the input and output for a speech-generation model like WaveNet?

I am currently studying this model speech generation known as WaveNet model by Google. https://arxiv.org/pdf/1609.03499.pdf using the linked original paper and this implementation. I find the model ...
user avatar
  • 123
1 vote
2 answers
29 views

Inconsistency on A. Graves' original Connectionist Temporal Classification (CTC) paper?

Here's the link to the paper. For the forward-backward algorithm, they introduce $\alpha_t(s)$ as a definition in eq. (5). Then they give a recursive formula for it (eq. (6)), and its initialization. ...
user avatar
  • 111
0 votes
1 answer
20 views

Split text into phrases of a person and an operator

I have about 5,000 texts without punctuation marks. Each text is a conversation between an operator and a person. For example: "Hello hello how can I help you how can I find out how much money I have ...
user avatar
  • 47
1 vote
1 answer
97 views

Conv1D specify output chanels in tensorflow 2.1

Hello I'm trying to implement the "Tacotron Towards end to end" paper and in the Encoder CBHG - a Conv1D bank of K=16, conv-k-128-ReLU “conv-k-c-ReLU” - denotes 1-D ...
user avatar
0 votes
1 answer
35 views

Detection of anomal data in the text

I work with texts where there is a dialogue between two people (a client and a call center employee, the beginning and end of each person’s phrase is not defined). My goal is to classify texts in ...
user avatar
  • 47
1 vote
1 answer
42 views

Why gaussian assumption in GMM-HMM ASR?

I am reading a book entitled Speech and Language Processing by Daniel Jurafsky and James. For Acoustic vector, I would like to know why Gaussian assumption is made? I searched over the internet and ...
user avatar
2 votes
2 answers
196 views

Creating pronunciation dictionary for ASR

I am working on ASR(automatic speech recoginition) on Somali data as master thesis and now I am stuck with how to create a phonetics or pronunciation dictionary for it. I searched over net and could ...
user avatar
2 votes
1 answer
88 views

GMM in speech recoginition using HMM-GMM

I am trying to solve/understand ASR using HMM-GMM. At the abstract level i do understand what's happening but I did not understand how GMM fits into it. My data has 5K hours of speech from single ...
user avatar
3 votes
1 answer
189 views

ASR on low dataset

I am doing an ASR(automatic speech recognition) as master thesis on low key dataset. Voice and text data is labelled. There are around 4000 phrases and around 5 hours speech. I don't have background ...
user avatar
1 vote
2 answers
116 views

where to start in natural language processing for a language

My native language is a regional language and few people speak it. I have some assignements in a machine learning course and i was thinking about doing some natural languge processing on my native ...
user avatar
0 votes
2 answers
313 views

Training an acoustic model for a speech-to-text engine

What are the steps for training an acoustic model? The format of the data (the audio) includes its length and other characteristics. If anyone could provide a simple example of how to train an ...
user avatar
  • 1,934
0 votes
1 answer
42 views

What to use in setting up a Speech to Text engine in production?

So i have the task to study the feasability of setting up a Speech-To-Text engine in a production environnement, and i've been researching on this topic, so I tried Google's Speech-To-Text API and ...
user avatar
  • 1,934
1 vote
1 answer
135 views

Segment 5-7 min audio into sentence wise audio clips for creating speech recognition dataset

I am trying to create a speech recognition dataset, especially for Indian Accents. I am taking from colleagues to build this. Daily I send an article link and ask them to record and upload it to ...
user avatar
1 vote
1 answer
69 views

Can I create a good Speech Recognition Engine while having millions of recorded conversations?

I have at my disposal millions of wav files containing recorded conversations between employees and clients, i'm doing some research on the possibility of creating a good speech recognition engine. I'...
user avatar
  • 1,934
0 votes
1 answer
44 views

Opensource Speech Recognition Library that is secure and trained on large data [closed]

For all those who are working on developing a chatbot/assistant and care about the privacy of users consuming the speech recognition library, can you suggest an open souce library which is trained on ...
user avatar
  • 113
4 votes
1 answer
67 views

Why are observation probabilities modelled as Gaussian distributions in HMM?

HMM is a statistical model with unobserved (i.e. hidden) states used for recognition algorithms (speech, handwriting, gesture, ...). What distinguishes DHMM form CHMM is the transition probability ...
user avatar
2 votes
0 answers
53 views

Representing output label for character level speec recognition using RNN

I saw this tutorial on generating text using LSTM. In this tutorial the author trained the network by taking 100 previous characters as input and the next character as the output label. I am ...
user avatar
2 votes
1 answer
400 views

Validation loss is less than training loss by 5 units. How this result is interpreted?

Iam training a Keras model for end-to-end speech recognition. I have my own dataset of speech containing about 400 wave files. Text transcriptions is also given as input. Model summary is: ...
user avatar
4 votes
2 answers
10k views

How to convert a mel spectrogram to log-scaled mel spectrogram

I was reading this paper on environmental noise discrimination using Convolution Neural Networks and wanted to reproduce their results. They convert WAV files into log-scaled mel spectrograms. How do ...
user avatar
  • 222
0 votes
1 answer
333 views

How can I create a space in IBM Cloud?

I want to try the IBM speech to text API. I created an IBM cloud account and went to https://console.bluemix.net/catalog/services/speech-to-text. I see the error message: Please create a space in ...
user avatar