Two Semi-Supervised Training Approaches for Automated Text Recognition

Abstract In the paper, we present EVALD applications (Evaluator of Discourse) for automated essay scoring. EVALD is the first tool of this type for Czech. It evaluates texts written by both native and non-native speakers of Czech. We describe first the history and the present in the automatic essay scoring, which is illustrated by examples of systems for other languages, mainly for English. Then we focus on the methodology of creating the EVALD applications and describe datasets used for testing as well as supervised training that EVALD builds on. Furthermore, we analyze in detail a sample of newly acquired language data – texts written by non-native speakers reaching the threshold level of the Czech language acquisition required e.g. for the permanent residence in the Czech Republic – and we focus on linguistic differences between the available text levels. We present the feature set used by EVALD and – based on the analysis – we extend it with new spelling features. Finally, we evaluate the overall performance of various variants of EVALD and provide the analysis of collected results.

Download Full-text

Deep Learning Techniques for Naskh and Nastalique Writing Style Text Recognition

International Journal of Computer Sciences and Engineering ◽

10.26438/ijcse/v7i4.7076 ◽

2019 ◽

Vol 7 (4) ◽

pp. 70-76

Author(s):

Shanky Goel ◽

Gurpreet Singh Lehal

Keyword(s):

Deep Learning ◽

Text Recognition ◽

Writing Style ◽

Learning Techniques

Download Full-text

The effect of proprioceptive training on postural control in people with diabetes: A randomized clinical trial comparing delivery at home, under supervision, or no training

Clinical Rehabilitation ◽

10.1177/0269215521989016 ◽

2021 ◽

pp. 026921552198901

Author(s):

Nathalia Cristina de Souza Borges ◽

Ariane Hidalgo Mansano Pletsch ◽

Mariana Barbosa Buzato ◽

Natalia Akemi Yamada Terada ◽

Fernanda Maria Ferreira da Cruz ◽

...

Keyword(s):

Type 2 Diabetes ◽

Postural Control ◽

Force Plate ◽

Outcome Variable ◽

Control Group ◽

Functional Tests ◽

Supervised Training ◽

Proprioceptive Training ◽

Home Training

Objective: Analyze postural control in the bipedal position as well as during gait and functional tests in patients with type 2 diabetes mellitus after supervised and unsupervised proprioceptive training. Design: A three-group randomized controlled trial. Setting: Physiotherapeutic Resources Lab, Department of Health Sciences, Ribeirão Preto Medical School, University of São Paulo Subjects: Eighty patients with type 2 diabetes allocated to three groups: control, home training, and supervised training. Interventions: The supervised and home training groups performed two weekly sessions of proprioceptive exercises for 12 weeks. The control group was not submitted to any of treatment. Main measures: Bipedal balance, gait, and performance on functional tests were evaluated before and after 12 weeks using the Balance Evaluation Systems Test (BESTest) and the force plate. Results: No significant improvements were found regarding postural control, gait, or performance on the functional tests, as evidenced by the inter-group comparisons of the total BESTest score [control: 90.7 (81.5–92.6); home training: 85.2 (77.8–90.3); supervised training: 88.4 (82.6–91.4), P > 0.05] as well as the tests performed on the force plate ( P > 0.05). The clinical effect size of the proposed intervention was less than 0.2, demonstrating no effect for the main outcome variable evaluated by the “Sensory Orientation” item of the BESTest and by the mCTSIB (pressure plate). Conclusions: The proposed proprioceptive training did not lead to improvements in postural control in patients with type 2 diabetes with no clinical signs of diabetic distal polyneuropathy when analyzed using the BESTest clinical evaluation and a force plate. Trial registration: NCT01861392 (clinicaltrials.gov).

Download Full-text