AI Paper Atlas

Research map / Label-efficient and robust learning / Knowledge distillation

Deep Ensemble Learning by Diverse Knowledge Distillation for Fine-Grained Object Classification

Naoki Okamoto, Tsubasa Hirakawa, Takayoshi Yamashita, Hironobu Fujiyoshi

ECCV 2022 · Label-efficient and robust learning › Knowledge distillation · approach: Rethinking distillation objectives · 13 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
optimized diverse-KD graph instead of independent training (2 nodes, ABN)ensemble of ABN · Stanford Dogsensemble accuracy (%): 70.90 -> 73.86Improved
optimized diverse-KD graph instead of DML (4 nodes, ABN)ensemble of ABN · Stanford Dogsensemble accuracy (%): 72.87 -> 74.16Improved
optimized graph instead of independent training (5 nodes, ResNet-18)ensemble of ResNet-18 · Stanford Dogsensemble accuracy (%): 69.47 -> 71.35Improved
DML instead of independent training (ResNet-18, 5 nodes)ensemble of ResNet-18 · Stanford Dogsensemble accuracy (%): 69.47 -> 69.94No clear effect
ONE multi-branch instead of independent networks (ABN, 5 branches)ensemble of ABN · Stanford Dogsensemble accuracy (%): 72.32 -> 70.89Hurt
reuse graph optimized on Stanford Dogs for CUB-200-2011 training2-node graph · CUB-200-2011ensemble accuracy (%): 65.26 -> 72.06Improved
reuse graph optimized on Stanford Dogs for CIFAR-100 training2-node graph · CIFAR-100ensemble accuracy (%): 73.16 -> 72.19Hurt
optimize graph on the same dataset for CIFAR-1002-node graph · CIFAR-100ensemble accuracy (%): 73.16 -> 74.18Improved
graph optimized on CUB-200-2011 itself rather than Stanford Dogs (ablation)2-node graph · CUB-200-2011ensemble accuracy (%): 72.06 -> 69.81Hurt
use CIFAR-100 graph on CIFAR-10 instead of Stanford Dogs graph (ablation)2-node graph · CIFAR-10ensemble accuracy (%): 93.87 -> 94.37Improved
distill from ensemble of optimized 2-node graph instead of DML teacherResNet-18 student · Stanford Dogsstudent accuracy (%): 69.68 -> 72.94Improved
larger divergence between networks' output distributionsensemble of two ResNets · Stanford Carsensemble gain correlation: 0.499Improved

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