# Deep Structured Output Learning for Unconstrained Text Recognition

arXiv: https://arxiv.org/abs/1412.5903 · HTML: https://mlsanity.com/paper/1412.5903/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 18 Dec 2014.

## Built on by (most cited first, 13 of 13)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Read Like Humans: Autonomous, Bidirectional and Iterative Language Modeling for Scene Text Recognition](https://mlsanity.com/paper/2103.06495/) (2021)
- [ABINet++: Autonomous, Bidirectional and Iterative Language Modeling for Scene Text Spotting](https://mlsanity.com/paper/2211.10578/) (2022)
- [Visual Semantics Allow for Textual Reasoning Better in Scene Text Recognition](https://mlsanity.com/paper/2112.12916/) (2021)
- [Towards the Unseen: Iterative Text Recognition by Distilling from Errors](https://mlsanity.com/paper/2107.12081/) (2021)
- [SGBANet: Semantic GAN and Balanced Attention Network for Arbitrarily Oriented Scene Text Recognition](https://mlsanity.com/paper/2207.10256/) (2022)
- [Convolutional Neural Networks with Gated Recurrent Connections](https://mlsanity.com/paper/2106.02859/) (2021)
- [Visual-Semantic Transformer for Scene Text Recognition](https://mlsanity.com/paper/2112.00948/) (2021)
- [Parallel Scale-wise Attention Network for Effective Scene Text Recognition](https://mlsanity.com/paper/2104.12076/) (2021)
- [Masked Vision-Language Transformers for Scene Text Recognition](https://mlsanity.com/paper/2211.04785/) (2022)
- [Scene Text recognition with Full Normalization](https://mlsanity.com/paper/2109.01034/) (2021)
- [SAFL: A Self-Attention Scene Text Recognizer with Focal Loss](https://mlsanity.com/paper/2201.00132/) (2022)
- [Scene Text Recognition with Image-Text Matching-guided Dictionary](https://mlsanity.com/paper/2305.04524/) (2023)

## Abstract (arXiv)

We develop a representation suitable for the unconstrained recognition of words in natural images: the general case of no fixed lexicon and unknown length. To this end we propose a convolutional neural network (CNN) based architecture which incorporates a Conditional Random Field (CRF) graphical model, taking the whole word image as a single input. The unaries of the CRF are provided by a CNN that predicts characters at each position of the output, while higher order terms are provided by another CNN that detects the presence of N-grams. We show that this entire model (CRF, character predictor, N-gram predictor) can be jointly optimised by back-propagating the structured output loss, essentially requiring the system to perform multi-task learning, and training uses purely synthetically generated data. The resulting model is a more accurate system on standard real-world text recognition benchmarks than character prediction alone, setting a benchmark for systems that have not been trained on a particular lexicon. In addition, our model achieves state-of-the-art accuracy in lexicon-constrained scenarios, without being specifically modelled for constrained recognition. To test the generalisation of our model, we also perform experiments with random alpha-numeric strings to evaluate the method when no visual language model is applicable.