Towards Unsupervised Learning for Handwriting Recognition

Kozielski, Michał; Nuhn, Malte; Doetsch, Patrick; Ney, Hermann

doi:10.1109/icfhr.2014.98

Cited by 6 publications

(2 citation statements)

References 16 publications

(20 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[46] learn mappings from hidden-states of an HMM with their transition probabilities initialised with conditional bi-gram distributions. [40] propose an iterative scheme for bootstrapping predictions for learning HMMs models, and recognise handwritten text. However, their approach is limited to (1) word images, (2) fixed lexicon (≈44K words) to facilitate exhaustive tree search, whereas, our method is applicable to full text strings, does not require a pre-defined lexicon of words.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Learning to Read by Spelling

Gupta

Vedaldi

Zisserman

2018

Proceedings of the 11th Indian Conference on Computer Vision, Graphics and Image Processing

View full text Add to dashboard Cite

tttttttttttttttttttttttttttttttttttttttttttrssssss ttttttt ny nytt nr nrttttttt ny ntt nrttttttt zzzz bcorote to whol th ticunthss tidio tiostolonzzzz trougfht to ferr oy disectins it has dicomered brought to view by dissection it was discovered Tr a i n i n g iterations Figure 1: Text recognition from unaligned data. We present a method for recognising text in images without using any labelled data. This is achieved by learning to align the statistics of the predicted text strings, against the statistics of valid text strings sampled from a corpus. The figure above visualises the transcriptions as various characters are learnt through the training iterations. The model first learns the concept of {space}, and hence, learns to segment the string into words; followed by common words like {to, it}, and only later learns to correctly map the less frequent characters like {v, w}. The last transcription also corresponds to the ground-truth (punctuations are not modelled). The colour bar on the right indicates the accuracy (darker means higher accuracy). ABSTRACTThis 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.

show abstract

Section: Related Workmentioning

confidence: 99%

“…Although earlier works use low-order, namely uni/bi-gram statistics for alignment [40,46], higher-order n-grams could be more informative. In this experiment we examine the impact of the length of the training text-sequences on convergence.…”

Section: Effect Of Text Length On Convergencementioning

confidence: 99%