AI Paper Atlas

Research map / Video / Video segmentation and tracking

Smoothing Slot Attention Iterations and Recurrences

Rongzhen Zhao, Wenyan Yang, Juho Kannala, Joni Pajarinen

ICML 2026 · Video › Video segmentation and tracking · approach: Spatio-temporal predictive modeling

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 query preheater (Transformer decoder block) with preheat loss (ablation)SmoothSA-i · COCO ablationARI + ARIfg: 56.9 -> 68.3Improved
use Transformer decoder block instead of Slot Attention module as preheater (ablation)SmoothSA-i · COCO ablationARI + ARIfg: 63.3 -> 68.3Improved
switch cross-attention and self-attention order in preheater (ablation)SmoothSA-i · COCO ablationARI + ARIfg: 49.6 -> 68.3Improved
stop gradient on preheated queries (ablation)SmoothSA-i · COCO ablationARI + ARIfg: 67.5 -> 68.3Improved
preheating loss weight 100 instead of 10 (ablation)SmoothSA-i · COCO ablationARI + ARIfg: 59.7 -> 68.3Improved
preheating loss weight 100 instead of 200 (ablation)SmoothSA-i · COCO ablationARI + ARIfg: 67.4 -> 68.3Improved
share SA weights between first and non-first transforms instead of separate (ablation)SmoothSA-v · YTVIS ablationARI + ARIfg: 52.3 -> 68.3Improved
use 3 first-frame iterations instead of 1 (unconditional video) (ablation)SmoothSA-v · YTVIS unconditional video ablationARI + ARIfg: 97.4 -> 105.6Improved
use 1 non-first-frame iteration instead of 3 (unconditional video) (ablation)SmoothSA-v · YTVIS unconditional video ablationARI + ARIfg: 103.4 -> 105.6Improved
use 3+1 iterations instead of 1+1 (conditional video) (ablation)SmoothSA-v · MOVi-C conditional video ablationARI + ARIfg: 133.9 -> 136.3Improved
use 3+1 iterations instead of 3+3 (conditional video) (ablation)SmoothSA-v · MOVi-C conditional video ablationARI + ARIfg: 132.7 -> 136.3Improved
replace DIAS SA with SmoothSA-iDIAS · COCO object discoverymIoU: 30.2 -> 31.8Improved

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