Vu Trong Phung's Audio Novels by Kaggle | Other

About this Dataset

Vu Trong Phung's Audio Novels

Audio files of novels and stories by Vu Trong Phung for Vietnamese TTS training

By Thông Nguyễn (From Huggingface) [source]

About this dataset

This dataset, named Vietnamese Text-to-Speech Dataset: Novels and Stories by Vu Trong Phung, is a collection of audio files featuring novels and stories written by the renowned Vietnamese author Vu Trong Phung. The purpose of this dataset, specifically the train.csv file, is to provide audio data for training a Vietnamese Text-to-Speech (TTS) model.

The train.csv file contains the necessary audio files required for training the TTS model. These audio files are sourced from Vu Trong Phung's novels and stories, ensuring a diverse range of linguistic patterns and tones representative of his literary works. By using this dataset, researchers and developers can create an accurate Vietnamese TTS system capable of converting written text into natural-sounding speech that preserves the style and essence of Vu Trong Phung's writings.

With access to these insightful audio recordings derived from his novels and stories, this dataset enables users to delve into different narrative styles, character portrayals, emotional expressions, and unique storytelling techniques employed by Vu Trong Phung. Such detailed information enhances the quality and authenticity of trained TTS models dedicated to reproducing his literary works in spoken form.

By leveraging this extensive collection as a training resource for Vietnamese TTS technology development projects or research endeavors in speech synthesis systems with a focus on Vu Trong Phung's literature, researchers can achieve more accurate prosody prediction while maintaining linguistic style consistency throughout their synthesized speech outputs. This contributes not only to technological advancements but also aids in preserving and promoting appreciation for classical Vietnamese literature through innovative applications like real-time audiobooks or virtual storytelling experiences.

Considered an invaluable repository within the field of natural language processing (NLP), this comprehensive collection provides an opportunity for scientists and enthusiasts alike to explore linguistic nuances embedded within these auditory renditions carefully curated from notable literary works authored by Vu Trong Phung. The availability of such rich information empowers NLP practitioners to enhance their TTS models' performance and elevate the overall user experience when interacting with synthesized Vietnamese speech generated from Vu Trong Phung's novels and stories

How to use the dataset

Dataset Overview

The dataset consists of audio files (.wav format) containing recordings of the novels and stories by Vu Trong Phung. These audio files are included in the audio folder.

File Descriptions

audio: This folder contains the audio files of the novels and stories written by Vu Trong Phung. Each audio file represents a different text passage or section from these literary works.

train.csv: This file provides information about each audio data entry in the dataset. It includes details such as metadata about each recording, including annotations, transcript text, duration, etc.

Dataset Usage

To make use of this dataset effectively, you can follow these steps:

Extract both the audio folder and train.csv file to your desired directory on your local machine or server.

Analyze and preprocess your data: You can start exploring the train.csv file to gain an understanding of its structure and content. Use this information for any necessary preprocessing steps on your audio data before training a TTS model.

Train your TTS model: Utilizing popular frameworks like Tensorflow or PyTorch, you can build a Vietnamese TTS model using this preprocessed data from Vu Trong Phung's novels as input features and corresponding transcripts as target labels.

Evaluate and fine-tune: Evaluate your trained model using appropriate evaluation metrics for speech synthesis tasks such as Mean Squared Error (MSE), Mel Cepstral Distortion (MCD), or Naturalness and Intelligibility scores. Make any necessary adjustments or fine-tuning to improve the performance of your TTS model.

Generate synthetic speech: With a trained and optimized TTS model, you can input new Vietnamese text passages and generate corresponding synthesized speech output. This could be useful for various applications such as audiobook production, voice assistants, or text-to-speech conversion services.

Remember to always acknowledge this dataset and provide appropriate credits to its authors when using it in your work.

Conclusion

Research Ideas

Training a Vietnamese Text-to-Speech (TTS) model: This dataset provides audio files of novels and stories by Vu Trong Phung, which can be used to train a TTS model for generating audio output from Vietnamese text.

Linguistic analysis: Researchers or linguists can use this dataset to analyze the linguistic features and patterns in the works of Vu Trong Phung. They can study the pronunciation, intonation, and other linguistic aspects of Vietnamese language as represented in the audio files.

Historical and cultural research: The novels and stories by Vu Trong Phung are significant works in Vietnamese literature. Researchers or historians interested in studying the historical and cultural context of that era can utilize this dataset to better understand the author's writing style, themes, and narratives through analyzing the audio files

Acknowledgements

If you use this dataset in your research, please credit the original authors.
Data Source

License

License: CC0 1.0 Universal (CC0 1.0) - Public Domain Dedication
No Copyright - You can copy, modify, distribute and perform the work, even for commercial purposes, all without asking permission. See Other Information.

Columns

File: train.csv

Column name	Description
audio	The file names or paths of the audio files. (String)

Acknowledgements

If you use this dataset in your research, please credit the original authors.
If you use this dataset in your research, please credit Thông Nguyễn (From Huggingface).

Tables

Train

@kaggle.thedevastator_vu_trong_phung_s_audio_novels.train

2.85 KB
2 rows
2 columns


CREATE TABLE train (
  "a" VARCHAR,
  "dio" VARCHAR
);