AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Drawing and Recognizing Chinese Characters with Recurrent Neural Network

Xu-Yao Zhang, Fei Yin, Yan‐Ming Zhang, Cheng‐Lin Liu, Yoshua Bengio

TPAMI 2017 · Architectures and efficiency › RNNs and normalization · approach: Efficient recurrent architectures · 372 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 LSTM units with GRU units (1 layer)bidirectional RNN · ICDAR-2013test accuracy %: 97.67 -> 97.71No clear effect
replace LSTM with GRU units (2 layers)bidirectional RNN · ICDAR-2013test accuracy %: 97.70 -> 97.76No clear effect
replace LSTM with GRU unitsbidirectional RNN · ICDAR-2013test time per character (ms): 0.3792 -> 0.2949Improved
replace LSTM with GRU units (3 layers)bidirectional RNN · ICDAR-2013test accuracy %: 97.80 -> 97.77No clear effect
replace LSTM with GRU unitsbidirectional RNN · ICDAR-2013memory: 11.00MB -> 9.06MBImproved
stack two bidirectional layers instead of one (GRU)bidirectional GRU · ICDAR-2013test accuracy %: 97.71 -> 97.76Improved
stack three bidirectional layers instead of one (LSTM)bidirectional LSTM · ICDAR-2013test accuracy %: 97.67 -> 97.80Improved
stack three layers instead of one (LSTM)bidirectional LSTM · ICDAR-2013training time (h): 95.16 -> 216.75Hurt
predict from a single sub-sequence from sequential dropout instead of full sequence (ablation)NET4 · ICDAR-2013test accuracy %: 97.76 -> 96.54Hurt
ensemble 30 sub-sequences from input dropout at test time (ablation)NET4 · ICDAR-2013test accuracy %: 97.76 -> 97.89Improved
ensemble 30 sub-sequences at test time (ablation)NET5 · ICDAR-2013test accuracy %: 97.80 -> 97.96Improved
average six separately trained networksNET1 to NET6 · ICDAR-2013test accuracy %: 97.76 -> 98.15Improved

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