Release Date: 11/19/2026
Format: TSV, WAV
Size: 2.73 GB
The Sylheti TTS Dataset is a 10-hour speech corpus featuring recordings from a male Sylheti speaker. The dataset is designed to support research and development in text-to-speech (TTS), automatic speech recognition (ASR), speech synthesis, and other speech and language technologies for Sylheti. The corpus provides valuable speech data for a relatively under-resourced language variety, helping researchers develop and evaluate speech technologies while contributing to the preservation and digital representation of Sylheti. The recordings can be used for building speech models, studying pronunciation and phonetic characteristics, and developing language resources for future Sylheti NLP and speech applications. This dataset is particularly useful for researchers, developers, and organizations working on low-resource speech technology, multilingual AI, language documentation, and computational linguistics.
Pricing details
Your purchase is the license to the raw data. Once purchased, you're responsible for storing and using this dataset.
Your purchase includes a license to the data, paid directly to the dataset vendor, and a 5% ($50.00) platform fee paid to MDC.
Read More About The PoliciesThe user/buyer agrees to MirasAI Data License terms.
The user/buyer agrees to MirasAI Data License terms.
Intended Use
The Sylheti TTS Dataset is intended for research and development in text-to-speech (TTS), automatic speech recognition (ASR), speech synthesis, pronunciation modeling, and other speech and language technologies.
Sylheti is a language variety spoken primarily in the Sylhet region of northeastern Bangladesh, as well as in parts of Assam and Tripura in India, and by large diaspora communities abroad (notably a significant population in the United Kingdom). It has distinctive phonology, vocabulary, and grammar that set it apart from Standard Bangla, and many linguists consider it a distinct language within the Bengali-Assamese language continuum rather than a simple dialect, making it a valuable target for speech technology and language documentation efforts.
অ, আ, ই, ঈ, উ, ঊ, ঋ, এ, ঐ, ও, ঔ, ক, খ, গ, ঘ, ঙ, চ, ছ, জ, ঝ, ঞ, ট, ঠ, ড, ঢ, ণ, ত, থ, দ, ধ, ন, প, ফ, ব, ভ, ম, য, র, ল, শ, ষ, স, হ, ড়, ঢ়, য়, ং, ঃ, ঁ।
| Field | Value |
|---|---|
| Speaker ID | 01 |
| Speaker Age | 21 |
| Speaker Region | Bangladesh |
| Speaker Gender | Male |
| Field | Value |
|---|---|
| Total Recordings | 3,173 |
| Total Duration | 10 hours |
| File Type | WAV (Studio Quality) |
| Bit Depth | 16-bit PCM |
| Sampling Rate | 48 kHz |
| Channels | Mono |
| Background Noise | Low (clean/studio-quality profile) |
| Signal-to-Noise Ratio (SNR) | High |
| Reverberation Time (RT60) | Estimated ~0.18–0.25 seconds |
| Domain | Number of Clips |
|---|---|
| TTS Dataset | 3,173 |
| Total Clips | 3,173 |
| Total Files (incl. TXT files) | 3,173 WAV files |