AI Paper Atlas

Research map / Architectures and efficiency / Quantization

Differentiable Soft Quantization: Bridging Full-Precision and Low-Bit Neural Networks

Ruihao Gong, Xianglong Liu, Shenghu Jiang, Tianxiang Li, Peng Hu, Jiazhen Lin, Fengwei Yu, Junjie Yan

ICCV 2019 · Architectures and efficiency › Quantization · approach: Mixed-precision quantization · 499 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
add DSQ with fixed alpha (ablation)2-bit uniform quantization · CIFAR-10 ResNet-20, 2-bit uniformtop-1 accuracy (%): 86.63 -> 86.95Improved
learn alpha during training (evolution training) (ablation)DSQ 2-bit · CIFAR-10 ResNet-20, 2-bit uniformtop-1 accuracy (%): 86.95 -> 87.25Improved
learn clipping values l and u jointly with alpha (ablation)DSQ 2-bit · CIFAR-10 ResNet-20, 2-bit uniformtop-1 accuracy (%): 87.25 -> 88.44Improved
add DSQ to PACT (ablation)PACT 2-bit · CIFAR-10 ResNet-20top-1 accuracy (%): 88.24 -> 90.11Improved
replace sign binarization with binary DSQ (alpha 0.2) (ablation)binarization 1/1 · CIFAR-10 ResNet-20accuracy (%): 82.46 -> 83.80Improved
use piecewise DSQ instead of binary DSQ (ablation)binarization 1/1 · CIFAR-10 ResNet-20accuracy (%): 83.80 -> 84.11Improved
use binary DSQ instead of plain binary (ablation)binarization 1/32 · CIFAR-10 ResNet-20accuracy (%): 90.11 -> 90.24Improved
use piecewise DSQ instead of binary DSQ (ablation)binarization 1/32 · CIFAR-10 ResNet-20accuracy (%): 90.24 -> 90.03Hurt
use DSQ instead of DoReFa-Net1-bit W/A 1/1 · CIFAR-10 ResNet-20accuracy (%): 79.30 -> 84.11Improved
use DSQ instead of PACT at 4-bitMobileNetV2 · ImageNet valaccuracy (%): 61.40 -> 64.80Improved
use DSQ instead of LQ-Net at 3-bitResNet-34 · ImageNet valaccuracy (%): 71.90 -> 72.54Improved
use DSQ 2-bit kernels instead of NCNN 8-bitResNet-18 inference · Raspberry Pi 3Btime (ms): 935.51 -> 551.22Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 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