AI Paper Atlas

Research map / Video / Video segmentation and tracking

Video Matting via Consistency-Regularized Graph Neural Networks

Tiantian Wang, Sifei Liu, Yapeng Tian, Kai Li, Ming-Hsuan Yang

ICCV 2021 · Video › Video segmentation and tracking · approach: Learned optical flow estimation · 29 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
add GNN inter-frame aggregation to image-level baseline (ablation)CRGNN video matting · Real human matting datasetMSE: 10.21 -> 9.480Improved
add GNN inter-frame aggregation to image-level baseline (ablation)CRGNN video matting · Real human matting datasetMESSDdt: 15.31 -> 13.45Improved
add consistency regularization (ablation)CRGNN video matting · Real human matting datasetMSE: 9.480 -> 9.260Improved
add consistency regularization (ablation)CRGNN video matting · Real human matting datasetSAD: 78.38 -> 73.21Improved
add consistency regularization (ablation)CRGNN video matting · Real human matting datasetMESSDdt: 13.45 -> 12.69Improved
add adversarial discriminator (ablation)CRGNN video matting · Real human matting datasetMSE: 9.260 -> 9.223Improved
add adversarial discriminator (ablation)CRGNN video matting · Real human matting datasetConnectivity: 60.75 -> 58.49Improved
add adversarial discriminator (ablation)CRGNN video matting · Real human matting datasetSAD: 73.21 -> 73.49No clear effect
use deformable alignment instead of non-local aggregation (ablation)CRGNN video matting · Real human matting datasetMSE: 9.954 -> 9.224Improved
use deformable alignment instead of non-local aggregation (ablation)CRGNN video matting · Real human matting datasetSAD: 89.45 -> 73.50Improved
increase test-time nodes from 3 to 5 (ablation)CRGNN video matting · Real human matting datasetMSE: 9.223 -> 9.230No clear effect
increase test-time nodes from 3 to 7 (ablation)CRGNN video matting · Real human matting datasetMSE: 9.223 -> 9.228No clear effect

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