AI Paper Atlas

Research map / 3D vision / Face and shape

Learning Canonical 3D Object Representation for Fine-Grained Recognition

Sunghun Joung, Seungryong Kim, Minsu Kim, Ig-Jae Kim, Kwanghoon Sohn

ICCV 2021 · 3D vision › Face and shape · approach: Primitive-based shape abstraction · 15 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 4-D positional encoding on the canonical warped feature (ablation)Ours w/o fs · CUB-Birdsaccuracy: 83.7 -> 86.8Improved
add 4-D positional encoding on the canonical warped feature (ablation)Ours w/o fs · Stanford-Carsaccuracy: 85.5 -> 93.2Improved
add shape encoder using 3D shape deformation (ablation)Ours w/o fs · CUB-Birdsaccuracy: 86.8 -> 88.4Improved
add shape encoder using 3D shape deformation (ablation)Ours w/o fs · Stanford-Carsaccuracy: 93.2 -> 94.7Improved
use 4-D sinusoidal coordinates instead of 2-D coordinates for positional encoding (ablation)Ours · CUB-Birdsaccuracy: 85.2 -> 86.8Improved
use 4-D sinusoidal coordinates instead of 2-D coordinates for positional encoding (ablation)Ours · Stanford-Carsaccuracy: 89.8 -> 93.2Improved
use 3 FC layers instead of 0 in the shape encoder (ablation)Ours · CUB-Birdsaccuracy: 86.8 -> 88.4Improved
use 4 FC layers instead of 3 in the shape encoder (ablation)Ours · CUB-Birdsaccuracy: 88.4 -> 87.9Hurt
use 4 FC layers instead of 3 in the shape encoder (ablation)Ours · Stanford-Carsaccuracy: 94.7 -> 94.4Hurt
add positional encoding to the appearance feature (ablation)Ours w/o PE, fs · CUB-Birds 3D reconstructionPCK: 51.5 -> 51.9Improved
add shape encoder (ablation)Ours w/o fs · CUB-Birds 3D reconstructionmask IoU: 0.732 -> 0.737Improved
replace 2D spatial deformation modules (VTN) with canonical 3D representationVTN · Stanford-Carsaccuracy: 82.7 -> 85.5Improved

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