The dataset is currently empty. Upload or create new data files. Then, you will be able to explore them in the Dataset Viewer.

Description

Burmese Scripted Monologue Smartphone speech dataset, collected from monologue based on given prompts. Transcribed with text content. Our dataset was collected from extensive and diversify speakers, geographicly speaking, enhancing model performance in real and complex tasks.

For more details, please refer to the link: https://www.nexdata.ai/datasets/speechrecog/1817?source=huggingface

Specifications

Format

16kHz, 16bit, uncompressed wav, mono channel

Recording environment

quiet indoor environment, without echo

Recording content (read speech)

News and general textss

Speaker

About 100 people.

Device

Android mobile phone, iPhone

Language

Burmese

Transcription content

text

Accuracy rate

Word Accuracy Rate (WAR) 98%

Application scenarios

speech recognition, voiceprint recognition

Licensing Information

Commercial License

Downloads last month
13