MatAnyone 2 (CVPR 2026 Highlight)
Source: https://github.com/pq-yang/matanyone2
Successor of pq-yang/MatAnyone.
Structured fields
{
"archived": false,
"checked_at": "2026-10-17T00:00:00.000Z",
"usefulness": "Video matting scaled with a learned quality evaluator (CVPR 2026 Highlight); adds CLI, uv install and Hugging Face support. Current line for high-quality alpha from a mask prompt.",
"source_type": "repository",
"availability": "available",
"source_owner": "Peiqing Yang et al. (S-Lab, NTU)",
"next_check_at": "2026-10-24T00:00:00.000Z",
"known_limitations": [
"Non-commercial license per README."
],
"upstream_updated_at": "2026-07-28T03:04:18.000Z"
}Evidence claims
- [upstream_claim] GitHub description: [CVPR 2026 Highlight] MatAnyone 2: Scaling Video Matting via a Learned Quality Evaluator; updates list 2026.03 CLI/uv/Hugging Face support and inference code release. · locator: README updates; repository description source
- [sourced_fact] README notes MatAnyone2Kit, a third-party Core ML/Swift port for Apple platforms (2026.06). · locator: README, Update source
- [sourced_fact] README License: NTU S-Lab License 1.0 (non-commercial). · locator: README, License source
- [upstream_claim] Research review verdict at export: reviewed (as of 2026-10-17). Recorded by the research pipeline, not a hub review outcome; evidence.review stays unreviewed until an independent reviewer records one.
Rights
| component | license | hosting | redistribution | modification | training | Evidence and notes |
|---|---|---|---|---|---|---|
| code | LicenseRef-NTU-S-Lab-1.0 | restricted | restricted | restricted | unknown | S-Lab License 1.0 per README; the sibling MatAnyone LICENSE text (read) limits use to non-commercial purposes. This repo's LICENSE file link was not found at the expected path; license taken from its README. |
| model_weights | unknown | unknown | unknown | unknown | unknown | Weights terms not verified. |