Datasets:
Update README with landmark annotation image
Browse files
README.md
CHANGED
|
@@ -70,6 +70,17 @@ The `landmarks.parquet` file contains:
|
|
| 70 |
| Right Eyebrow | 43-47 |
|
| 71 |
| Mouth | 48-67 |
|
| 72 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 73 |
## Joining with Main Metadata
|
| 74 |
|
| 75 |
The main metadata (`welsh-speech-dataset`) is **sequence-level** (one row per speaker-phrase).
|
|
|
|
| 70 |
| Right Eyebrow | 43-47 |
|
| 71 |
| Mouth | 48-67 |
|
| 72 |
|
| 73 |
+
## Preview
|
| 74 |
+
|
| 75 |
+
<table>
|
| 76 |
+
<tr>
|
| 77 |
+
<td><img src="example-landmark-anno.png" width="500"></td>
|
| 78 |
+
</tr>
|
| 79 |
+
<tr>
|
| 80 |
+
<td align="center">Landmark Annotations</td>
|
| 81 |
+
</tr>
|
| 82 |
+
</table>
|
| 83 |
+
|
| 84 |
## Joining with Main Metadata
|
| 85 |
|
| 86 |
The main metadata (`welsh-speech-dataset`) is **sequence-level** (one row per speaker-phrase).
|