The DOI matches its Crossref record, with a small difference in title or year.
Shortest path segmentation: a method for training a neural network to recognize character strings
Abstract
The authors describe a method which combines dynamic programming and a neural network recognizer for segmenting and recognizing character strings. The method selects the optimal consistent combination of cuts from a set of candidate cuts generated using heuristics. The optimal segmentation is found by representing the image, the candidate segments, and their scores as a graph in which the shortest path corresponds to the optimal interpretation. The scores are given by neural net outputs for each segment. A significant advantage of the method is that the labor required to segment images manually is eliminated. The system was trained on approximately 7000 unsegmented handwritten zip codes provided by the United States Postal Service. The system has achieved a per-zip-code raw recognition rate of 81% on a 2368 handwritten zip-code test set.>
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“A now classical method for segmentation and recognition is called HOS [68], [69].Its main advantages over other approaches to segmentation are that it avoids making hard decisions about the segmentation by taking a large number of different segmentations into consideration.”
From Gradient-based learning applied to document recognition (1998) · cites this paper as [69]
Citations
Cites
No references could be resolved yet.
Research works printed but not linked: 11
- martin, 1991, Recognizing Overlapping Hand-Printed Characters by Centered-Object Integrated Segmentation and Recognition, Proc Neural Inform Processing Syst Conf
- 10.1109/2.144441 · DOI · Paper page
- 10.1109/5.18626 · DOI · Paper page
- lecolinet, 1989, A New System for Automatic Segmentation and Recognition of Unconstrained Handwritten ZIP Codes, Scandinavian Conference on Image Analysis
- 10.1109/PROC.1973.9030 · DOI
- keeler, 1990, Integrated Segmentation and Recognition of Hand-Printed Numerals, Proc Neural Inform Processing Syst Conf
- fenrich, 1990, United States Postal Service - Advanced Technology Conference, 1
- 10.1109/ICPR.1990.119325 · DOI
- shridhar, 1990, Session Chairman, Session on Hand Written Digit Recognition, United States Postal Service - Advanced Technology Conference
- bridle, 1989, Neuro-computing Algorithms Architectures and Applications
- fenrich, 1991, Segmentation of Automatically Located Handwritten Words, International Workshop on Frontiers in Handwriting Recognition
Cited by
Sorted by citations
Gradient-based learning applied to document recognition
Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998) · Proceedings of the IEEE · doi:10.1109/5.726791 · Cited by 58,805
Self-citation
Metadata
- Authors
- Burges, C. J. C., Matan, O., Le Cun, Y., Denker, J. S., Jackel, L. D., Stenard, C. E., Nohl, C. R., & Ben, J. I.
- Affiliations
- AT&T (United States) (US)
- ORCID
- Jackel, L.D.
- Venue
- [Proceedings 1992] IJCNN International Joint Conference on Neural Networks
- Publisher
- IEEE
- Published
- 2003-01-02
- Volume
- 3
- Pages
- 165-172
- ISSN
- 2161-4407 (ISSN-L)
- OpenAlex
- W2132793646
- Cited by
- 37
- References
- 0
- Citations by year
- 2020: 1 · 2018: 1 · 2016: 1 · 2014: 1 · 2013: 3
- FWCI
- 0.99
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- International Joint Conference on Neural Networks (IJCNN) (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Handwritten Text Recognition Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Vehicle License Plate Recognition
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Keywords
- dynamic programming, graph-based segmentation, handwritten character recognition
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
