AI Paper Atlas

Research map / Detection and segmentation / CNN features

DenseNets Reloaded: Paradigm Shift Beyond ResNets and ViTs

DongHyun Kim, Byeongho Heo, Dongyoon Han

ECCV 2024 · Detection and segmentation › CNN features · approach: Convolution-transformer backbones · 70 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
use concatenation instead of additive shortcuts in random networks (ablation)RandNet (Tiny-ImageNet) · Tiny-ImageNet random networkstop-1 accuracy: 55.9 -> 56.8 (averaged over 15k networks)Improved
add modernized feature-mixer blocks (LN, depthwise conv, fewer norms/acts, kernel 7) (ablation)wide shallow DenseNet · ImageNet-1K valtop-1 accuracy: 79.5 -> 80.4Improved
widen (GR 32 to 120) and make DenseNet-201 shallower (ablation)DenseNet-201 · ImageNet-1K valtop-1 accuracy / training speed: 79.7 -> 79.5 acc; train memory 3.9 -> 3.2 GBMixed
decouple expansion ratio from growth rate and halve GR (ablation)RDNet · ImageNet-1K valtop-1 accuracy: 80.4 -> 80.8Improved
add transition layer every three blocks and raise GR (ablation)RDNet · ImageNet-1K valtop-1 accuracy: 80.8 -> 82.3Improved
add patchification stem (ablation)RDNet · ImageNet-1K valtop-1 accuracy: 82.3 -> 82.4No clear effect
add channel re-scaling (layer-scale plus squeeze-excitation) (ablation)RDNet · ImageNet-1K valtop-1 accuracy: 82.6 -> 82.8Improved
add more activations in the block (ablation)RDNet-T · ImageNet-1K valtop-1 accuracy: 82.8 -> 81.9Hurt
use stochastic depth ratio 0.15 (ablation)RDNet-T · ImageNet-1K valtop-1 accuracy: 81.6 -> 82.8Improved
use variable growth rate per stage instead of uniform GR 120 (ablation)RDNet-T · ImageNet-1K valtop-1 accuracy / FLOPs: 83.0 (8.9 GFLOPs) -> 82.8 (5.0 GFLOPs)Mixed
transition ratio 0.5 versus higher ratios (ablation)RDNet-T · ImageNet-1K valtop-1 accuracy: 0.7: 82.3 -> 0.5: 82.8Improved
use RDNet-B instead of ConvNeXt-B in CLIP trainingCLIP on CC3M/CC12M/RedCaps · ImageNet-1K zero-shottop-1 accuracy: 51.2 -> 54.1Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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