Sangbeom Lim, Seoung Wug Oh, Heeji Yoon, Seungryong Kim, Joon-Young Lee · 2026 · arXiv
DOI: 10.48550/arXiv.2610.26857 · open-access PDF
Mating video berbasis difusi baru-baru ini muncul sebagai pendekatan terkemuka, memanfaatkan priors difusi video dari pra-latihan skala internet untuk mencapai generalisasi nol-shot yang kuat dari data sintetis terbatas. Namun, perhitungan berat tulang belakang difusi ini membatasi inferensi ke potongan pendek dari hanya 6 hingga 24 bingkai, sehingga bahkan video biasa meliputi beberapa potongan. Karena setiap potongan dibatasi secara independen dari panduan yang kasar, metode yang ada gagal mempertahankan konsistensi di seluruh batas potongan, menghasilkan tepi berkedip dan transparansi yang tidak berkelanjutan.
Machine translation (Meta M2M100); the English original follows.
Diffusion-based video matting has recently emerged as a leading approach, leveraging video diffusion priors from internet-scale pretraining to achieve strong zero-shot generalization from limited synthetic data. However, the heavy compute of these diffusion backbones limits inference to a short chunk of just 6 to 24 frames, so even ordinary videos span multiple chunks. Because each chunk is matted independently from coarse mask guidance, existing methods fail to maintain consistency across chunk boundaries, producing flickering edges and discontinuous transparency. We address this with a mask-
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.26857, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange