Update README.md
Browse files
README.md
CHANGED
|
@@ -29,6 +29,11 @@ We created this dataset to support physically grounded video generation, moving
|
|
| 29 |
|
| 30 |
By identifying where current models break physical rules, we aim to enable more reliable and trustworthy generative video systems.
|
| 31 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 32 |
## 🛡️ Content Filtering
|
| 33 |
In this open-source version, we apply filtering procedures to reduce privacy and intellectual property risks:
|
| 34 |
- 🚫 Videos with identifiable human faces are excluded
|
|
|
|
| 29 |
|
| 30 |
By identifying where current models break physical rules, we aim to enable more reliable and trustworthy generative video systems.
|
| 31 |
|
| 32 |
+
## 🆕 Caption Release Update
|
| 33 |
+
Due to strong community demand and multiple email inquiries, we have decided to also open-source the caption associated with each video generation.
|
| 34 |
+
|
| 35 |
+
To reduce copyright risk in our initial video release, we removed the first five frames from each generated video, since the WISA dataset was crawled from online content. For users interested in the image prompt, we recommend using the first remaining frame of each uploaded video as a reasonable proxy for the initial image prompt used for the corresponding generation.
|
| 36 |
+
|
| 37 |
## 🛡️ Content Filtering
|
| 38 |
In this open-source version, we apply filtering procedures to reduce privacy and intellectual property risks:
|
| 39 |
- 🚫 Videos with identifiable human faces are excluded
|