Character-Based Handwritten Text Transcription with Attention Networks

Abstract

The paper approaches the task of handwritten text recognition (HTR) with attentional encoder-decoder networks trained on sequences of characters, rather than words. We experiment on lines of text from popular handwriting datasets and compare different activation functions for the attention mechanism used for aligning image pixels and target characters. We find that softmax attention focuses heavily on individual characters, while sigmoid attention focuses on multiple characters at each step of the decoding. When the sequence alignment is one-to-one, softmax attention is able to learn a more precise alignment at each step of the decoding, whereas the alignment generated by sigmoid attention is much less precise. When a linear function is used to obtain attention weights, the model predicts a character by looking at the entire sequence of characters and performs poorly because it lacks a precise alignment between the source and target. Future research may explore HTR in natural scene images, since the model is capable of transcribing handwritten text without the need for producing segmentations or bounding boxes of text in images.

References in corpus (11)

• Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation
• Convolutional Sequence to Sequence Learning
• Attention-Based Models for Speech Recognition
• Grammar as a Foreign Language
• On Using Very Large Target Vocabulary for Neural Machine Translation
• Unsupervised Adaptation for Synthetic-to-Real Handwritten Word Recognition
• Text and Style Conditioned GAN for Generation of Offline Handwriting Lines
• Have convolutions already made recurrence obsolete for unconstrained handwritten text recognition ?
• OrigamiNet: Weakly-Supervised, Segmentation-Free, One-Step, Full Page Text Recognition by learning to unfold
• CTCModel: a Keras Model for Connectionist Temporal Classification
• WSRNet: Joint Spotting and Recognition of Handwritten Words

Cited by in corpus (1)

• Understanding attention-based encoder-decoder networks: a case study with chess scoresheet recognition