Learning to Read by Spelling: Towards Unsupervised Text Recognition
arXiv:1809.08675
Abstract
This work presents a method for visual text recognition without using any paired supervisory data. We formulate the text recognition task as one of aligning the conditional distribution of strings predicted from given text images, with lexically valid strings sampled from target corpora. This enables fully automated, and unsupervised learning from just line-level text-images, and unpaired text-string samples, obviating the need for large aligned datasets. We present detailed analysis for various aspects of the proposed method, namely - (1) impact of the length of training sequences on convergence, (2) relation between character frequencies and the order in which they are learnt, (3) generalisation ability of our recognition network to inputs of arbitrary lengths, and (4) impact of varying the text corpus on recognition accuracy. Finally, we demonstrate excellent text recognition accuracy on both synthetically generated text images, and scanned images of real printed books, using no labelled training examples.
References in corpus (15)
- Sequence to Sequence Learning with Neural Networks
- TensorFlow: A system for large-scale machine learning
- Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation
- Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks
- Synthetic Data and Artificial Neural Networks for Natural Scene Text Recognition
- Unsupervised Image-to-Image Translation Networks
- Unsupervised Machine Translation Using Monolingual Corpora Only
- Synthetic Data for Text Localisation in Natural Images
- End-to-End Text Recognition with Hybrid HMM Maxout Models
- Unsupervised Neural Machine Translation
- Deep Structured Output Learning for Unconstrained Text Recognition
- An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognition
- Recursive Recurrent Nets with Attention Modeling for OCR in the Wild
- Unsupervised Cipher Cracking Using Discrete GANs
- Towards Principled Unsupervised Learning