AI Paper Atlas

Research map / 3D vision / 3D generation

Safe Multi-View Deep Classification

Wei Liu, Yufei Chen, Xiaodong Yue, Changqing Zhang, Shaorong Xie

AAAI 2023 · 3D vision › 3D generation · approach: 2D-to-3D pretrained representations · 33 citations (OpenAlex)

Read the paper

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
Use evidence-level fusion with uncertainty and KL-divergence losses instead of the strongest compared baselineMulti-view classification · HMDBaccuracy (%): 74.98±1.02 -> 90.84±0.11Improved
Add uncertainty loss to evidence fusion without KL-divergence loss (ablation)SMDC · Caltech101accuracy (%): 96.00±0.12 -> 96.89±0.20Improved
Add KL-divergence loss to evidence fusion without uncertainty loss (ablation)SMDC · Caltech101accuracy (%): 96.00±0.12 -> 97.33±0.12Improved
Add uncertainty and KL-divergence losses to evidence fusion (ablation)SMDC · Caltech101accuracy (%): 96.00±0.12 -> 97.78±0.01Improved
Add view V3 to V1+V2SMDC · Handwrittenaccuracy (%): 85.25±0.04 -> 96.75±0.01Improved
Add view V4 to V1+V2+V3SMDC · Handwrittenaccuracy (%): 96.75±0.01 -> 97.50±0.02Improved
Add view V5 to V1+V2+V3+V4SMDC · Handwrittenaccuracy (%): 97.50±0.02 -> 98.75±0.02Improved
Add view V6 to V1+V2+V3+V4+V5SMDC · Handwrittenaccuracy (%): 98.75±0.02 -> 99.00±0.01Improved
Add a noisy view at noise ratio λ = 0.5 rather than using the pre-addition baselineSMDC · CUBaccuracy (%): 74.17±0.01 -> 78.31±0.01Improved
Add a noisy view at noise ratio λ = 0.5 rather than using the pre-addition baselineSMDC · Scene15accuracy (%): 56.96±1.23 -> 65.33±1.12Improved

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