AI Paper Atlas

Research map / Architectures and efficiency / Quantization

Effective Training of Convolutional Neural Networks With Low-Bitwidth Weights and Activations

Bohan Zhuang, Mingkui Tan, Jing Liu, Lingqiao Liu, Ian D. Reid, Chunhua Shen

TPAMI 2021 · Architectures and efficiency › Quantization · approach: Mixed-precision quantization · 42 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 two-stage training (TS) (ablation)2-bit ResNet-50 DoReFa · ImageNet val 2W2ATop-1 accuracy (%): 70.19 -> 70.92Improved
add progressive precision (PP) (ablation)2-bit ResNet-50 DoReFa · ImageNet val 2W2ATop-1 accuracy (%): 70.19 -> 70.78Improved
add TS and PP together (ablation)2-bit ResNet-50 DoReFa · ImageNet val 2W2ATop-1 accuracy (%): 70.19 -> 71.13Improved
add stochastic precision (SP)DoReFa-Net ResNet-50 · ImageNet valTop-1 accuracy (%): 70.19 -> 72.23Improved
add stochastic precision (SP)BiReal-Net ResNet-18 · ImageNet valTop-1 accuracy (%): 56.43 -> 58.81Improved
add stochastic precision (SP)GroupNet ResNet-18 · ImageNet valTop-1 accuracy (%): 64.82 -> 65.89Improved
add random weight/activation quantization to layerdrop (ablation)GroupNet + layerdrop · ImageNet valTop-1 accuracy (%): 65.7 -> 65.0Hurt
add random weight/activation quantization to blockdrop (ablation)GroupNet + blockdrop · ImageNet valTop-1 accuracy (%): 65.6 -> 65.9Improved
add joint knowledge distillation with same-size teacher2-bit ResNet-50 · ImageNet val 2W2ATop-1 accuracy (%): 70.19 -> 71.96Improved
jointly train teacher instead of freezing it (ablation)2-bit PreResNet-18 KD · ImageNet val 2W2ATop-1 accuracy (%): 65.09 -> 65.67Improved
use deeper PreResNet-50 teacher instead of PreResNet-34 (ablation)2-bit PreResNet-18 · ImageNet val 2W2ATop-1 accuracy (%): 65.82 -> 65.86No clear effect
add attention-transfer guidance to posterior distillation (ablation)2-bit ResNet-50 KD · ImageNet val 2W2ATop-1 accuracy (%): 71.40 -> 71.96Improved

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