AI Paper Atlas

Research map / Video / Video segmentation and tracking

Betrayed by Attention: A Simple yet Effective Approach for Self-supervised Video Object Segmentation

Shuangrui Ding, Rui Qian, Haohang Xu, Dahua Lin, Hongkai Xiong

ECCV 2024 · Video › Video segmentation and tracking · approach: Mask propagation for unsupervised video segmentation · 10 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 motion consistency loss on top of semantic loss with entropy weighting (ablation)BA · YTVIS-19FG-ARI: 30.1 -> 43.1 (without w), 44.3 with wImproved
add semantic consistency loss and entropy weighting to untrained block (ablation)BA · YTVIS-19FG-ARI: 23.4 -> 30.1Improved
add entropy-based importance weighting (ablation)BA · DAVIS-17 UnsupervisedFG-ARI: 38.1 -> 40.1Improved
cluster learned spatio-temporal attention instead of frame-wise DINO attention (ablation)BA · YTVIS-19FG-ARI: 30.1 -> 44.3Improved
cluster attention maps rather than output features (ablation)BA · YTVIS-19FG-ARI: 33.9 -> 44.3Improved
replace hierarchical clustering with K-means (8 centroids) (ablation)BA · YTVIS-19FG-ARI: 44.3 -> 42.5No clear effect
use temporal stride 4 instead of 1 in training (ablation)BA · YTVIS-19FG-ARI: 37.4 -> 44.3Improved
increase frames from 3 to 5 (ablation)BA · YTVIS-19FG-ARI: 44.3 -> 44.8No clear effect
use self-attention maps as motion representation instead of global correlation (ablation)BA · YTVIS-19FG-ARI: 39.1 -> 44.3Improved
use DINO ViT-S/8 instead of ViT-S/16 (ablation)BA · YTVIS-19FG-ARI: 42.5 -> 44.3Improved
swap DINO for DINOv2 backbone (ablation)BA · YTVIS-19FG-ARI: 44.3 -> 44.1 (DINOv2 ViT-S/14)No clear effect
sample only 10% of frames as keys at inference (ablation)BA · YTVIS-19FG-ARI: 44.3 -> 43.5, 2.3x speedupNo 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