AI Paper Atlas

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

Self-Supervised Real-to-Sim Scene Generation

Aayush Prakash, Shoubhik Debnath, Jean-Francois Lafleche, Eric Cameracci, Gavriel State, Stan Birchfield, Marc T. Law

ICCV 2021 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Pretext-task visual pre-training · 27 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 label alignment sigma^c,label (ablation)Sim2SG with appearance and prediction alignment · KITTI hardmAP@0.5 IoU: 0.246 -> 0.316Improved
add label alignment sigma^c,label (ablation)Sim2SG with appearance and prediction alignment · KITTI hardrelationship triplet Recall@50: 0.076 -> 0.139Improved
add appearance alignment sigma^a to label alignment (ablation)Sim2SG · KITTI hardmAP@0.5 IoU: 0.285 -> 0.311Improved
add appearance alignment sigma^a to label alignment (ablation)Sim2SG · KITTI hardrelationship triplet Recall@50: 0.104 -> 0.127Improved
add prediction alignment sigma^c,pred (ablation)Sim2SG with label and appearance alignment · KITTI hardrelationship triplet Recall@50: 0.127 -> 0.139Improved
add appearance and prediction alignment to label alignment (ablation)Sim2SG · KITTI hard, Pedestrian APAP@0.5 IoU: 0.262 -> 0.241Hurt
use Sim2SG instead of SDR synthetic datadetector trained on synthetic data · KITTI hardmAP@0.5 IoU: 0.234 -> 0.316Improved
use Sim2SG instead of Meta-Simscene generation · KITTI hardrelationship triplet Recall@50: 0.075 -> 0.139Improved
use GPA feature alignment instead of SDR onlydomain adaptation baseline · KITTI hardmAP@0.5 IoU: 0.234 -> 0.087Hurt
add appearance alignment sigma^a to label alignment (ablation)Sim2SG · CLEVR target domainrelationship triplet Recall@20: 0.493 -> 0.888Improved
add appearance alignment sigma^a in a content-gap-only setting (ablation)Sim2SG vs SDR · CLEVR (same appearance, different content)relationship triplet Recall@20: 0.760 -> 0.722Hurt
add label alignment sigma^c,label in a content-gap-only setting (ablation)Sim2SG vs SDR · CLEVR (same appearance, different content)relationship triplet Recall@20: 0.76 -> 0.996Improved

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