|
Download README.md from bobokkk/VideoSynthesis: direct link, hf CLI and curl.
- Browser
- Download file 1.32 kB
-
https://huggingface.co/datasets/bobokkk/VideoSynthesis/resolve/main/README.md
- Command line
-
hf download hf://datasets/bobokkk/VideoSynthesis/README.md
-
curl -L -o README.md https://huggingface.co/datasets/bobokkk/VideoSynthesis/resolve/main/README.md
1.32 kB
| language: | |
| - en | |
| tags: | |
| - synthetic | |
| - video | |
| - face-swapping | |
| pretty_name: VideoSynthesis | |
| # VideoSynthesis | |
| AI-generated face-swap clips from the representative-clip strategy. Legacy runs, | |
| full source videos, reference portraits, model weights, and credentials are not | |
| included. Each published clip is a finalized DreamID-V output upscaled with | |
| SeedVR2, with the driving clip's original audio retained. | |
| `representative/metadata.jsonl` indexes the current version of each sample. Video | |
| paths are repository-relative. Adjacent JSON sidecars preserve the per-version | |
| metadata. Files use content-addressed names; older versions may remain in Git | |
| history or the repository, so use the index rather than globbing all videos. | |
| Metadata identifies source IDs and time windows, not verified real-world names. | |
| The clean windows are selected by automated screening; this is not a guarantee | |
| that every output is artifact-free. Splits marked `unassigned` are not a | |
| validated identity-disjoint train/test split. Reused identities and source | |
| segments must be accounted for before training or evaluation. | |
| This upload does not grant additional rights to source media or model assets. | |
| Source permissions and applicable model licenses remain the uploader's | |
| responsibility. Synthetic outputs must not be represented as authentic footage. | |