AI Paper Atlas

Research map / Architectures and efficiency / Quantization

ReCU: Reviving the Dead Weights in Binary Neural Networks

Zihan Xu, Mingbao Lin, Jianzhuang Liu, Jie Chen, Ling Shao, Yue Gao, Yonghong Tian, Rongrong Ji

ICCV 2021 · Architectures and efficiency › Quantization · approach: Binary neural networks · 96 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
apply ReCU clamp with fixed tau=0.92 instead of no clamping (tau=1.00) (ablation)binary ResNet-18 · CIFAR-100, binary ResNet-18top-1 accuracy: 67.55 -> 68.47Improved
lower fixed tau to 0.82, the quantization-error minimum (ablation)binary ResNet-18 · CIFAR-100, binary ResNet-18top-1 accuracy: 68.47 -> 66.50Hurt
lower fixed tau to 0.78 (ablation)binary ResNet-18 · CIFAR-100, binary ResNet-18top-1 accuracy: 68.47 -> 59.09Hurt
replace fixed tau=0.92 with exponential scheduler 0.85/0.99 (ablation)ReCU · CIFAR-100, binary ResNet-18top-1 accuracy: 68.47 -> 68.69Improved
use end value 0.99 instead of 1.00 in scheduler (start 0.85) (ablation)ReCU scheduler · CIFAR-100, binary ResNet-18top-1 accuracy: 68.55 -> 68.69Improved
add generalized weight standardization with b*=sqrt(2)/2 (ablation)ReCU · CIFAR-100, binary ResNet-18top-1 accuracy: 66.13 -> 68.69Improved
raise b* from 0.2 to 2 (ablation)ReCU standardization · CIFAR-100, binary ResNet-18top-1 accuracy: 68.10 -> 69.02Improved
raise b* from 2 to 5 (ablation)ReCU standardization · CIFAR-100, binary ResNet-18top-1 accuracy: 69.02 -> 68.48Hurt
use ReCU vs vanilla BNN at 100 training epochsbinary ResNet-18 · CIFAR-100, 100 epochstop-1 accuracy: 52.1 -> 66.3Improved
use ReCU vs vanilla BNN at 600 training epochsbinary ResNet-18 · CIFAR-100, 600 epochstop-1 accuracy: 62.0 -> 69.1Improved
use ReCU instead of RBNNbinary ResNet-18 · ImageNettop-1 accuracy: 59.9 -> 61.0Improved
use ReCU instead of RBNNbinary ResNet-34 · ImageNettop-1 accuracy: 63.1 -> 65.1Improved

Later papers that compared with this method

8 later accepted papers report ReCU as a baseline in their results tables; 8 of them report a better number on most of the results they share with it.

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