On-Line and Off-Line Handwriting Recognition: A Comprehensive Survey
IEEE Transactions on Pattern Analysis and Machine Intelligence
The SMART Retrieval System—Experiments in Automatic Document Processing
The SMART Retrieval System—Experiments in Automatic Document Processing
ICML '06 Proceedings of the 23rd international conference on Machine learning
Matching ottoman words: an image retrieval approach to historical document indexing
Proceedings of the 6th ACM international conference on Image and video retrieval
Text search for medieval manuscript images
Pattern Recognition
A Novel Connectionist System for Unconstrained Handwriting Recognition
IEEE Transactions on Pattern Analysis and Machine Intelligence
Handwritten word-spotting using hidden Markov models and universal vocabularies
Pattern Recognition
Automatic Transcription of Handwritten Medieval Documents
VSMM '09 Proceedings of the 2009 15th International Conference on Virtual Systems and Multimedia
Semi-Supervised Learning
A novel word spotting algorithm using bidirectional long short-term memory neural networks
ANNPR'10 Proceedings of the 4th IAPR TC3 conference on Artificial Neural Networks in Pattern Recognition
Hi-index | 0.00 |
Keyword spotting is the task of retrieving all instances of a given keyword in a set of documents. In the current paper we consider the problem of keyword spotting in handwritten text. This is a difficult problem due to the great variety of different writing styles. Recently, learning based keyword spotting systems have been shown to outperform traditional approaches, at the cost of requiring large amounts of training data. The training data need to be manually labeled, which is tedious and time-consuming. In this paper we propose to exploit unlabeled data via semi-supervised learning to reduce the need for labeled data when training a keyword spotting system. We demonstrate, on historic as well as modern handwritten text, that the performance of a learning based keyword spotting system can be dramatically increased using this approach.