AI Paper Atlas

Research map / Architectures and efficiency / Quantization

BASQ: Branch-Wise Activation-Clipping Search Quantization for Sub-4-Bit Neural Networks

Han-Byul Kim, Eunhyeok Park, Sungjoo Yoo

ECCV 2022 · Architectures and efficiency › Quantization · approach: Mixed-precision quantization · 6 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
use BASQ activation quantization instead of LSQ (ablation)2-bit ResNet-20 · CIFAR10top-1 accuracy (%): 89.40 -> 90.21Improved
use BASQ activation quantization instead of LSQ (ablation)2-bit MobileNet-v2 · CIFAR10top-1 accuracy (%): 89.62 -> 90.47Improved
add new building block with dedicated BatchNorm (ablation)2-bit ResNet-20 · CIFAR10top-1 accuracy (%): 89.40 -> 89.72Improved
add new building block with dedicated BatchNorm (ablation)2-bit MobileNet-v2 · CIFAR10top-1 accuracy (%): 89.62 -> 89.74Improved
use FracBNN block instead of basic block (ablation)2-bit ResNet-20 · CIFAR10top-1 accuracy (%): 89.40 -> 89.39No clear effect
use FracBNN block instead of basic block (ablation)2-bit MobileNet-v2 · CIFAR10top-1 accuracy (%): 89.62 -> 89.32Hurt
add all three Flexconn skip connections with new block (ablation)2-bit MobileNet-v2 · CIFAR10top-1 accuracy (%): 89.74 -> 90.94Improved
share convolution across branches instead of private (ablation)BASQ 2-bit ResNet-20 · CIFAR10top-1 accuracy (%): 89.90 -> 90.21Improved
search L2 decay weight instead of random selection (ablation)2-bit MobileNet-v2 · ImageNettop-1 accuracy (%): [0.08%, 57.76%] range of random archs vs 64.71Improved
random L2 decay selection (ablation)2-bit ResNet-18 · ImageNettop-1 accuracy (%): [67.98%, 68.53%]No clear effect
train LSQ with BASQ 2-phase schedule, compared to BASQ (ablation)2-bit MobileNet-v2 · ImageNettop-1 accuracy (%): 50.0 -> 64.7Improved
train LSQ with BASQ 2-phase schedule, compared to BASQ (ablation)4-bit MobileNet-v2 · ImageNettop-1 accuracy (%): 70.2 -> 72.0Improved

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