AI Paper Atlas

Research map / Graph learning / Knowledge and temporal graphs

DCPB: Deformable Convolution Based on the Poincare Ball for Top-view Fisheye Cameras

Xuan Wei, Zhidan Ran, Xiaobo Lu

ICCV 2023 · Graph learning › Knowledge and temporal graphs · approach: Temporal graph neural architectures

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 regular conv with modulated deformable convolution (ablation)UNet · THEODORE testmIoU (%): 90.85 -> 91.97Improved
compute offsets and modulation scalars with a GCN in Euclidean space (ablation)DC · THEODORE testmIoU (%): 91.97 -> 92.57Improved
move the GCN from Euclidean space to the Poincare ball (ablation)DC+GCN · THEODORE testmIoU (%): 92.57 -> 92.74Improved
add geodesic distances to Poincare hyperplanes as offset parameters (ablation)DC+GCNP · THEODORE testmIoU (%): 92.74 -> 93.05Improved
apply DCPB at the high-level UP1 stage rather than a low-level stage (ablation)UNet-DCPB · THEODORE testmIoU (%): 92.69 (Up3) -> 93.05 (Up1)Improved
add DCPB layersUNet · THEODORE testFPS: 13.8 -> 13.4Hurt
use DCPB instead of FisheyeHDK tangent-space hyperbolic deformable convUNet · THEODORE testmIoU (%): 92.07 -> 93.05Improved

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