Update README.md
Browse files
README.md
CHANGED
|
@@ -51,7 +51,7 @@ sourced from the [NPTEL](https://nptel.ac.in/) platform. The original TIE datase
|
|
| 51 |
with each lecture averaging about 50 minutes. These lectures cover a wide range of technical subjects and capture diverse linguistic features characteristic of Indian
|
| 52 |
English.
|
| 53 |
|
| 54 |
-
The TIE_shorts version (~ 70 hours) was created to facilitate efficient training and usage in speech processing tasks by providing shorter audio samples. In TIE_shorts,
|
| 55 |
consecutive audio snippets from the original dataset were merged based on timestamps, with a condition that the final merged audio should not exceed 30 seconds in duration.
|
| 56 |
This process results in 25–30 second audio clips, each accompanied by a corresponding ground-truth transcript. This approach retains the linguistic diversity of the original
|
| 57 |
dataset while significantly reducing the size and complexity, making TIE_shorts ideal for Automatic Speech Recognition (ASR) and other speech-to-text applications.
|
|
|
|
| 51 |
with each lecture averaging about 50 minutes. These lectures cover a wide range of technical subjects and capture diverse linguistic features characteristic of Indian
|
| 52 |
English.
|
| 53 |
|
| 54 |
+
The TIE_shorts version (~ 70 hours audio and 600K ground-truth tokens) was created to facilitate efficient training and usage in speech processing tasks by providing shorter audio samples. In TIE_shorts,
|
| 55 |
consecutive audio snippets from the original dataset were merged based on timestamps, with a condition that the final merged audio should not exceed 30 seconds in duration.
|
| 56 |
This process results in 25–30 second audio clips, each accompanied by a corresponding ground-truth transcript. This approach retains the linguistic diversity of the original
|
| 57 |
dataset while significantly reducing the size and complexity, making TIE_shorts ideal for Automatic Speech Recognition (ASR) and other speech-to-text applications.
|