Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -46,7 +46,7 @@ This dataset contains 850 hours processed webinar audio segments with correspond
|
|
46 |
- `speaker`: Speaker identifier (e.g., "SPEAKER_00")
|
47 |
|
48 |
#### Quality Metrics
|
49 |
-
- `
|
50 |
- `emos_1`, `emos_2`, `emos_3`: EMOS quality scores
|
51 |
- `noise_confidence`: Noise detection confidence
|
52 |
|
@@ -54,7 +54,6 @@ This dataset contains 850 hours processed webinar audio segments with correspond
|
|
54 |
|
55 |
|
56 |
#### Segment Structure
|
57 |
-
- `segment_type`: Type of segment ("merged" or "individual")
|
58 |
- `num_sentences`: Number of sentences (for merged segments)
|
59 |
- `original_segments`: Original subsegments data (for merged segments)
|
60 |
|
@@ -77,7 +76,7 @@ The dataset consists of webinar recordings that have been processed and segmente
|
|
77 |
|
78 |
### Loading the Dataset
|
79 |
|
80 |
-
Load all
|
81 |
|
82 |
## Additional Information
|
83 |
|
|
|
46 |
- `speaker`: Speaker identifier (e.g., "SPEAKER_00")
|
47 |
|
48 |
#### Quality Metrics
|
49 |
+
- `emos_overall`: It's not utmos, it's EMOS overall quality score
|
50 |
- `emos_1`, `emos_2`, `emos_3`: EMOS quality scores
|
51 |
- `noise_confidence`: Noise detection confidence
|
52 |
|
|
|
54 |
|
55 |
|
56 |
#### Segment Structure
|
|
|
57 |
- `num_sentences`: Number of sentences (for merged segments)
|
58 |
- `original_segments`: Original subsegments data (for merged segments)
|
59 |
|
|
|
76 |
|
77 |
### Loading the Dataset
|
78 |
|
79 |
+
Load all files and unpack using ```cat webinars_stripped_archive.tar.aa webinars_stripped_archive.tar.ab webinars_stripped_archive.tar.ac webinars_stripped_archive.tar.ad webinars_stripped_archive.tar.ae webinars_stripped_archive.tar.af webinars_stripped_archive.tar.ag > webinars_stripped_archive.tar && tar -xf webinars_stripped_archive.tar```
|
80 |
|
81 |
## Additional Information
|
82 |
|