Upload README.md
Browse files
README.md
ADDED
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
|
2 |
+
|
3 |
+
# Auto-ACD
|
4 |
+
Auto-ACD is a large-scale, high-quality, audio-language dataset, building on the prior of robust audio-visual correspondence in existing video datasets, VGGSound and AudioSet.
|
5 |
+
|
6 |
+
|
7 |
+
- **Homepage:** https://auto-acd.github.io/
|
8 |
+
- **Paper:**
|
9 |
+
- **Github:** https://github.com/LoieSun/Auto-ACD
|
10 |
+
|
11 |
+
## Analysis
|
12 |
+
|
13 |
+
![](src/analysis.png)
|
14 |
+
|
15 |
+
Auto-ACD</strong>, comprising over <strong>1.9M </strong> audio-text pairs.
|
16 |
+
As shown in figure, The text descriptions in Auto-ACD contain <strong>long texts (18 words)</strong> and <strong>diverse vocabularies (23K)</strong>, and provide information about the <strong>surrounding auditory environment</strong>(data point with <strong>shadow</strong>) in which sounds take place.
|
17 |
+
|
18 |
+
## Download
|
19 |
+
|
20 |
+
We provide a csv file. For each data pairs, we provide YouTube URLs and generated caption. Each line in the csv file has columns defined by here.
|
21 |
+
|
22 |
+
```
|
23 |
+
# YouTube ID, caption
|
24 |
+
```
|
25 |
+
|
26 |
+
## Dataset Preview
|
27 |
+
|
28 |
+
![](src/samples.png)
|
29 |
+
|