AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Deep Learning for Fixed Model Reuse

Yang Yang, De-Chuan Zhan, Ying Fan, Yuan Jiang, Zhi-Hua Zhou

AAAI 2017 · Architectures and efficiency › CNN design · approach: Compact CNN architectures · 50 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
parallel structure with knockdown instead of sequential target replacing (SQR) (ablation)FMR · WIKIaccuracy (%): 42.7 -> 43.3Improved
parallel structure with knockdown instead of SQR (ablation)FMR · Flickr8kaccuracy (%): 72.6 -> 74.8Improved
parallel structure with knockdown instead of SQR (ablation)FMR · HMDB51accuracy (%): 62.9 -> 68.9Improved
parallel structure with knockdown instead of SQR (ablation)FMR · UCF101accuracy (%): 85.3 -> 91.6Improved
parallel structure with knockdown instead of SQR (ablation)FMR · UCF50accuracy (%): 91.1 -> 92.4Improved
double features dropped per iteration from 52 to 104 (ablation)FMR · UCF101accuracy (%): 91.6 -> 92.0No clear effect
add FMR (reusing fixed text features) to CNN trained on images onlyCNN · WIKIaccuracy (%): 42.6 -> 43.3Improved
use FMR instead of MIFS fixed features at test timeFMR · UCF50accuracy (%): 94.4 -> 92.4Hurt
add knockdown to FMR training (ablation)CNN · UCF101accuracy (%)Improved

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