AI Paper Atlas

Research map / Architectures and efficiency / Quantization

PROFIT: A Novel Training Method for sub-4-bit MobileNet Models

Eunhyeok Park, Sungjoo Yoo

ECCV 2020 · Architectures and efficiency › Quantization · approach: Mixed-precision quantization · 76 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 PROFIT (with TS and PG) (ablation)PACT, 4-bit MobileNet-v3 · ImageNet val, 4-bit MobileNet-v3 ablationtop-1 accuracy [%]: 70.160 -> 72.948Improved
add PROFIT (with TS and PG) (ablation)DuQ, 4-bit MobileNet-v3 · ImageNet val, 4-bit MobileNet-v3 ablationtop-1 accuracy [%]: 71.466 -> 73.812Improved
add PROFIT alone (ablation)DuQ, 4-bit MobileNet-v3 · ImageNet val, 4-bit MobileNet-v3 ablationtop-1 accuracy [%]: 69.504 -> 72.260Improved
use DuQ instead of PACT (ablation)4-bit MobileNet-v3 · ImageNet val, 4-bit MobileNet-v3 ablationtop-1 accuracy [%]: 67.978 -> 69.504Improved
use DuQ+PROFIT instead of PACT under TS and PG (ablation)4-bit MobileNet-v3 · ImageNet val, 4-bit MobileNet-v3 ablationtop-1 accuracy [%]: 70.160 -> 73.812Improved
add teacher-student distillation (ablation)DuQ, 4-bit MobileNet-v3 · ImageNet val, 4-bit MobileNet-v3 ablationtop-1 accuracy [%]: 69.504 -> 71.006Improved
add progressive quantization (ablation)DuQ+TS, 4-bit MobileNet-v3 · ImageNet val, 4-bit MobileNet-v3 ablationtop-1 accuracy [%]: 71.006 -> 71.466Improved
use full PROFIT instead of only the final normalization-layer stage (ablation)DuQ+TS+PG · ImageNet val, 4-bit MobileNet-v3 ablationtop-1 accuracy [%]: 72.826 -> 73.812Improved
use negative padding instead of zero padding with symmetric quantization (ablation)DuQ, 4-bit MobileNet-v3 · ImageNet val, 4-bit MobileNet-v3 ablationtop-1 accuracy [%]: 68.480 -> 69.504Improved
use negative padding instead of zero padding with non-negative quantization (ablation)DuQ, 4-bit MobileNet-v3 · ImageNet val, 4-bit MobileNet-v3 ablationtop-1 accuracy [%]: 68.724 -> 69.504Improved
use QIL activation quantization (ablation)4-bit MobileNet-v3 · ImageNet val, 4-bit MobileNet-v3 ablationtop-1 accuracy [%]: failHurt
add PROFIT to PACT at 3 bits (ablation)PACT, MobileNet-v3 3-bit · ImageNet val, 4-bit MobileNet-v3 ablation (3-bit)top-1 accuracy [%]: 57.086 -> 66.458Improved

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