AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Pure Transformer with Integrated Experts for Scene Text Recognition

Yew Lee Tan, Adams Wai-Kin Kong, Jung-Jae Kim

ECCV 2022 · Architectures and efficiency › CNN design · approach: Compact CNN architectures · 26 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 a pure ViT encoder instead of ResNet-45 plus transformer encoder (ablation)transformer STR model · 6 benchmarks weighted avg (7672)accuracy: 85.7 -> 90.9Improved
train experts with diverse patch resolutions and ground-truth orders instead of identical ones (ablation)PTIE · 6 benchmarks weighted avg (7672)accuracy: 91.0 -> 92.4Improved
integrate four experts in one model (PTIE-Diverse) instead of one 8x4 modelViT baseline · 6 benchmarks weighted avg (7672)accuracy: 90.9 -> 92.4Improved
replace four-model ensemble with PTIE-DiverseEnsemble-Diverse · 6 benchmarks weighted avg (7672)accuracy: 92.4 -> 92.4No clear effect
replace four-model ensemble with PTIE-DiverseEnsemble-Diverse · latencyinference time (ms): 202 -> 52Improved
use diverse instead of identical members in the ensemble (ablation)Ensemble · 6 benchmarks weighted avg (7672)accuracy: 92.1 -> 92.4Improved
add PTIE integrated experts to the baselineViT baseline · latencyinference time (ms): 50 -> 52No clear effect
use untied positional attention instead of vanilla positional encoding (ablation)PTIE · IC15 (2077)accuracy: 83.4 -> 84.3Improved
use untied positional attention instead of vanilla positional encoding (ablation)PTIE · IIIT5Kaccuracy: 96.7 -> 96.3Hurt
use untied positional attention instead of vanilla positional encoding (ablation)PTIE · weighted average over benchmarksaccuracy: 0.1% higherMixed
fine-tune with real data COCO-TextPTIE-Untied · SVT-Paccuracy: 90.1 -> 92.1Improved
fine-tune with real data COCO-TextPTIE-Untied · CUTE80accuracy: 91.7 -> 91.0Hurt

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