AI Paper Atlas

Research map / Video / Video segmentation and tracking

SIGMA: Sinkhorn-Guided Masked Video Modeling

Mohammadreza Salehi, Michael Dorkenwald, Fida Mohammad Thoker, Efstratios Gavves, Cees Snoek, Yuki M Asano

ECCV 2024 · Video › Video segmentation and tracking · approach: Self-supervised video representation learning · 7 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 pixel reconstruction targets with Sinkhorn-clustered learned features (MLP projection)VideoMAE · UCF-101 linear probetop-1 accuracy (%): 58.6 -> 73.8Improved
use frozen DINO as projection network (ablation)SIGMA · K400 linear probetop-1 accuracy (%): 30.7 -> 47.5Improved
use L2 loss on learned deep targets instead of Sinkhorn cluster prediction (ablation)SIGMA-MLP · mini SSv2 (ViT-S)top-1 accuracy (%): 60.3 -> 36.4Hurt
increase number of prototypes from 1000 to 4000 (ablation)SIGMA-MLP · mini SSv2 (ViT-S)top-1 accuracy (%): 59.7 -> 60.3Improved
increase prototypes from 4000 to 6000 (ablation)SIGMA-MLP · mini SSv2 (ViT-S)top-1 accuracy (%): 60.3 -> 60.1No clear effect
change projection MLP depth or width (ablation)SIGMA-MLP · mini SSv2 (ViT-S)top-1 accuracy (%): 59.7 -> 59.2 (shallower), 59.6 (wider)No clear effect
replace VideoMAE pixel targets with SIGMA-MLP in full finetuning on K400VideoMAE ViT-B · K400 full finetuningtop-1 accuracy (%): 80.0 -> 80.2Improved
use SIGMA-DINO versus SVT on multi-label CharadesSIGMA · SEVERE Charadesscore: 35.5 (SVT) vs 23.3Hurt
use SIGMA features for unsupervised video object segmentationVideoMAE · YTVOS overclustering ViT-BmIoU: 49.7 -> 60.4 (SIGMA-DINO)Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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