AI Paper Atlas

Research map / 3D vision / 3D generation

SceneGraphNet: Neural Message Passing for 3D Indoor Scene Augmentation

Yang Zhou, Zachary While, Evangelos Kalogerakis

ICCV 2019 · 3D vision › 3D generation · approach: Open-vocabulary 3D scene understanding · 85 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 dense co-occurrence edges to the scene graph (ablation)SceneGraphNet · SUNCG scene augmentationtop-1 accuracy: 60.0 -> 67.3Improved
use full graph with all relationships instead of tree structure (ablation)SceneGraphNet · SUNCG scene augmentationtop-1 accuracy: 61.1 -> 67.3Improved
add supporting/surrounding relationships to co-occurrence-only graph (ablation)SceneGraphNet · SUNCG scene augmentationtop-1 accuracy: 56.5 -> 67.3Improved
use GRU aggregation instead of summation (ablation)SceneGraphNet · SUNCG scene augmentationtop-1 accuracy: 57.7 -> 67.3Improved
use GRU aggregation instead of max-pooling (ablation)SceneGraphNet · SUNCG scene augmentationtop-1 accuracy: 63.1 -> 67.3Improved
use GRU aggregation instead of vanilla RNN (ablation)SceneGraphNet · SUNCG scene augmentationtop-1 accuracy: 64.8 -> 67.3Improved
add learned attention on messages (ablation)SceneGraphNet · SUNCG scene augmentationtop-1 accuracy: 60.3 -> 67.3Improved
use learned attention instead of distance-based weights (ablation)SceneGraphNet · SUNCG scene augmentationtop-1 accuracy: 63.8 -> 67.3Improved
run second message passing iteration instead of one (ablation)SceneGraphNet · SUNCG scene augmentationtop-5 accuracy: 72.7 -> 88.2Improved
run third message passing iteration instead of two (ablation)SceneGraphNet · SUNCG scene augmentationtop-5 accuracy: 88.2 -> 89.5Improved
combine SceneGraphNet posterior with MVCNN (ablation)MVCNN · SUNCG object recognitionaccuracy: 59.2 -> 72.2Improved
use SceneGraphNet instead of GRAINS for size prediction (ablation)scene augmentation · SUNCG size predictionsize error (cm): 38 -> 26Improved

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