AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Time Does Tell: Self-Supervised Time-Tuning of Dense Image Representations

Mohammadreza Salehi, Efstratios Gavves, Cees G.M. Snoek, Yuki M. Asano

ICCV 2023 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Pretext-task visual pre-training · 21 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
train on multiple frames with feature forwarding instead of single frames (ablation)TimeT · Pascal VOC K=500mIoU: 26.2 -> 42.8Improved
forward Sinkhorn-Knopp clustered features instead of raw logits (ablation)Feature-Forwarder · Pascal VOC K=500mIoU: 35.5 -> 42.8Improved
forward SK-clustered features instead of identity (static) forwarding (ablation)Feature-Forwarder · Pascal VOC K=500mIoU: 36.1 -> 42.8Improved
set time interval to 0.5s instead of 0.2s or 1.0s (ablation)TimeT · Pascal VOC K=500mIoU: 36.5 / 38.1 -> 42.8Improved
increase number of frames from 4 to 8 (ablation)TimeT · Pascal VOC K=500mIoU: 42.8 -> 42.3No clear effect
increase prototypes from 10 to 200 (ablation)TimeT · Pascal VOC K=500mIoU: 24.1 -> 42.8Improved
increase prototypes from 200 to 300 (ablation)TimeT · Pascal VOC K=500mIoU: 42.8 -> 42.3No clear effect
apply TimeT to different pretrained backbones (MSN, iBOT, DINO ViT-S/8, DINO ViT-B/16)TimeT · Pascal VOC K=500mIoU: 26.0 -> 48.3 (MSN)Improved
use feature forwarder instead of optical flow for mask propagation (ablation)Feature-Forwarder · DAVIS validationmIoU: 28.5 -> 40.3Improved
fine-tune image-based Leopart on video frames instead of imagesLeopart · Pascal VOC linear classificationmIoU: 68.0 -> 53.2Hurt
time-tuning DINO with video clustering loss versus DINO alone on video segmentationDINO · DAVIS clustering per-framemIoU: 30.2 -> 53.7Improved
TimeT trained on video versus image-trained Leopart on Pascal K=21TimeT · Pascal VOC K=21mIoU: 36.6 -> 34.5No 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