AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Shape Adaptor: A Learnable Resizing Module

Shikun Liu, Zhe Lin, Yilin Wang, Jianming Zhang, Federico Perazzi, Edward Johns

ECCV 2020 · Architectures and efficiency › CNN design · approach: Filter design and trained-filter analysis · 5 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
replace human-designed resizing layers with shape adaptorsMobileNetv2 · Birdsaccuracy (%): 60.37 -> 68.53Improved
replace human-designed resizing layers with shape adaptorsMobileNetv2 · Aircraftaccuracy (%): 77.64 -> 83.00Improved
replace human-designed resizing layers with shape adaptorsResNet-50 · CIFAR-100accuracy (%): 78.53 -> 80.29Improved
replace human-designed resizing layers with shape adaptorsResNet-50 · CIFAR-10accuracy (%): 95.50 -> 95.48No clear effect
replace human-designed resizing layers with shape adaptorsVGG-16 · ImageNetaccuracy (%): 73.92 -> 73.53Hurt
replace human-designed resizing layers with shape adaptorsResNet-50 · ImageNetaccuracy (%): 77.18 -> 78.74Improved
use learned shape during training instead of retraining with fixed final shape (Fix Final) (ablation)Shape Adaptor VGG-16 · CIFAR-100accuracy (%): 78.62 -> 79.16Improved
use learned shape instead of fixed largest shape (Fix Large) (ablation)Shape Adaptor VGG-16 · Aircraftaccuracy (%): 84.49 -> 86.95Improved
use 6 shape adaptors per heuristic instead of 5 (ablation)VGG-16 · Aircraftaccuracy (%): 84.80 -> 86.95Improved
use 4 shape adaptors per heuristic instead of 8 (ablation)VGG-16 · CIFAR-100accuracy (%): 78.16 -> 79.16Improved
initialise s(alpha) at 0.52 instead of heuristic 0.58 (ablation)Shape Adaptor VGG-16 · Aircraftaccuracy (%): 83.90 -> 86.95Improved
apply AutoSC shape compression to MobileNetv2MobileNetv2 · Aircraftaccuracy (%): 77.64 -> 78.95Improved

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