AI Paper Atlas

Research map / Architectures and efficiency / Quantization

Quantized Densely Connected U-Nets for Efficient Landmark Localization

Zhiqiang Tang, Xi Peng, Shijie Geng, Lingfei Wu, Shaoting Zhang, Dimitris Metaxas

ECCV 2018 · Architectures and efficiency › Quantization · approach: Post-training quantization of ViTs · 141 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
replace stacked U-Nets with order-1 densely connected U-Nets (16 U-Nets)stacked U-Nets · MPII validationparameters: 50.5M -> 15.9MImproved
replace stacked U-Nets with DU-Net (8 U-Nets)stacked U-Nets · MPII validationPCKh: 89.3 -> 89.5No clear effect
replace stacked U-Nets with DU-Net (4 U-Nets)stacked U-Nets · MPII validationPCKh: 88.3 -> 88.2No clear effect
add order-1 connectivity (order-1 vs order-0) (ablation)stacked U-Nets · MPII validationPCKhImproved
raise connectivity order from 1 to 7 (ablation)order-1 DU-Net(8) · MPII validation (epoch 150)PCKh: 87.1 -> 86.7Hurt
use memory-efficient shared-memory implementationDU-Net order-1 · 12GB GPU batch 16max trainable U-Nets: 9 -> 16Improved
add a second iterative pass (detection supervision)DU-Net(4) · 300-W full setNME: 4.01 -> 3.93Improved
add a second iterative pass (regression supervision)DU-Net(4) · 300-W full setNME: 3.34 -> 3.28Improved
use regression instead of detection supervisionDU-Net(4) · 300-W full setNME: 4.01 -> 3.34Improved
add iterative refinement to DU-Net(4) instead of doubling depth to DU-Net(8)DU-Net(4) · 300-W full setNME: 3.28 vs 3.26No clear effect
binarize weights without scaling factorDU-Net(4) · 300-W full setNME: 3.38 -> 3.75Hurt
add float scaling factor alpha to binarized weights (ablation)BW DU-Net(4) · 300-W full setNME: 3.75 -> 3.58Improved

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