AI Paper Atlas

Research map / Video / Video segmentation and tracking

BiMatting: Efficient Video Matting via Binarization

Haotong Qin, Lei Ke, Xudong Ma, Martin Danelljan, Yu-Wing Tai, Chi-Keung Tang, Xianglong Liu, Fisher Yu

NeurIPS 2023 · Video › Video segmentation and tracking · approach: Video restoration with temporal priors · 14 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace binarized MobileNetV3 encoder with SBB (ablation)binarized RVM · VM 512x288MAD: 28.49 -> 14.81Improved
add SAB sparse decoder alone to binarized RVM (ablation)binarized RVM · VM 512x288MAD: 28.49 -> 189.13Hurt
add SAB on top of SBB (ablation)RVM+SBB · VM 512x288MAD: 14.81 -> 12.82Improved
add SAB on top of SBB (FLOPs) (ablation)RVM+SBB · VM 512x288FLOPs (G): 0.57 -> 0.37Improved
add SBB+SAB to binarized RVM (dtSSD) (ablation)binarized RVM · VM 512x288dtSSD: 3.64 -> 2.69Improved
BiMatting vs RVM-ReAct at high resolutionRVM-ReAct · D646 2048x2048SAD: 57.38 -> 52.85Improved
BiMatting vs RVM-ReAct low-resRVM-ReAct · D646 512x512MAD: 56.41 -> 32.74Improved
BiMatting vs BGMv2 high-res AIMBGMv2 · AIM 2048x2048SAD: 45.76 vs 48.27Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy