Type

Conference Proceedings

Authors

Andy Way
Lambert Schomaker
Gideon Maillette de Buy Wenniger

Subjects

Computer Science

Topics
multi dimensional long short term memory machine learning fast deep learning variable length input handwriting recognition artificial intelligence deep learning example packing

No padding please: efficient neural handwriting recognition (2019)

Abstract Neural handwriting recognition (NHR) is the recognition of handwritten text with deep learning models, such as multi-dimensional long short-term memory (MDLSTM) re-current neural networks. Models with MDLSTM layers have achieved state-of-the art results on handwritten text recognition tasks. While multi-directional MDLSTM-layers have an unbeaten ability to capture the complete context in all directions, this strength limits the possibilities for parallelization, and therefore comes at a high computational cost.In this work we develop methods to create efficient MDLSTM-based models for NHR, particularly a method aimed at eliminating computation waste that results from padding. This proposed method, called example-packing, replaces wasteful stacking of padded examples with efficient tiling in a 2-dimensional grid.For word-based NHR this yields a speed improvement of factor6.6 over an already efficient baseline of minimal padding foreach batch separately. For line-based NHR the savings are more modest, but still significant.In addition to example-packing, we propose: 1) a technique to optimize parallelization for dynamic graph definition frameworks including PyTorch, using convolutions with grouping, 2) a method for parallelization across GPUs for variable-length example batches. All our techniques are thoroughly tested on our own PyTorch re-implementation of MDLSTM-based NHR models. A thorough evaluation on the IAM dataset shows that our models are performing similar to earlier implementations of state-of-theart models. Our efficient NHR model and some of the reusable techniques discussed with it offer ways to realize relatively efficient models for the omnipresent scenario of variable-length inputs in deep learning.
Collections Ireland -> Dublin City University -> Publication Type = Conference or Workshop Item
Ireland -> Dublin City University -> DCU Faculties and Centres = DCU Faculties and Schools: Faculty of Engineering and Computing: School of Computing
Ireland -> Dublin City University -> DCU Faculties and Centres = Research Initiatives and Centres: ADAPT
Ireland -> Dublin City University -> Subject = Computer Science: Artificial intelligence
Ireland -> Dublin City University -> Status = In Press
Ireland -> Dublin City University -> DCU Faculties and Centres = DCU Faculties and Schools: Faculty of Engineering and Computing
Ireland -> Dublin City University -> Subject = Computer Science: Machine learning

Full list of authors on original publication

Andy Way, Lambert Schomaker, Gideon Maillette de Buy Wenniger

Experts in our system

1
Andy Way
Dublin City University
Total Publications: 229
 
2
Gideon Maillette de Buy Wenniger
Dublin City University
Total Publications: 6