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.
Metadata
Item Type:
Conference or Workshop Item (Paper)
Event Type:
Conference
Refereed:
Yes
Additional Information:
This is a pre-publication of a paper which has been accepted at the
International Conference on Document Analysis and Recognition 2019
(ICDAR 2019, https://icdar2019.org/).
Uncontrolled Keywords:
variable length input; example-packing; multi-dimensional long short-term memory; handwriting recognition;deep learning; fast deep learning
European Union’s Horizon 2020 under the European Union’s Horizon 2020 research and innovthe Marie Skłodowska-Curie grant agreement No 713567., ADAPT Centre under the SFI Research Centres Programme (Grant 13/RC/2106).
ID Code:
23382
Deposited On:
03 Jul 2019 12:06 by
Gideon Maillette De buy
. Last Modified 17 Feb 2020 15:50