Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -12,7 +12,8 @@ language:
|
|
| 12 |
size_categories:
|
| 13 |
- 1K<n<10K
|
| 14 |
---
|
| 15 |
-
#
|
|
|
|
| 16 |
|
| 17 |
## 📌 Overview
|
| 18 |
**HumSpeechBlend** is a dataset designed to fine-tune **Voice Activity Detection (VAD) models** to distinguish between **humming** and actual speech. Current VAD models often misclassify humming as speech, leading to incorrect segmentation in speech processing tasks. This dataset provides a structured collection of humming audio interspersed with speech to help improve model accuracy.
|
|
|
|
| 12 |
size_categories:
|
| 13 |
- 1K<n<10K
|
| 14 |
---
|
| 15 |
+
# WORK IN PROGRESS
|
| 16 |
+
# [WIP]HumSpeechBlend Dataset: Humming vs Speech Detection
|
| 17 |
|
| 18 |
## 📌 Overview
|
| 19 |
**HumSpeechBlend** is a dataset designed to fine-tune **Voice Activity Detection (VAD) models** to distinguish between **humming** and actual speech. Current VAD models often misclassify humming as speech, leading to incorrect segmentation in speech processing tasks. This dataset provides a structured collection of humming audio interspersed with speech to help improve model accuracy.
|