Machine Learning Based Braille Transliteration of O dia Language

Abstract

Braille transliteration of natural languages is required for providing a better opportunity of learning and creating opportunities of ceceity people. It allows a bigger diaspora of non-blind teachers to have written communication with blind people. The present paper proposes a method of Braille transliteration of Handwritten and printed Odia characters automatically into Braille. The current work proposes a method of Braille transliteration of Handwritten Odia text with industry applicable accuracy. The method first preprocesses the text and then segments it into characters and then uses an SVM classifier trained on HOG features of Odia handwritten characters to predict characters and maps the predicted printable character to its corresponding Braille with a very good accuracy. The method can further be used with text to speech engines to help the blind students use this technique with refreshable Braille having audio facility to listen the same.

Detection of Touchscreen-Based Urdu Braille Characters Using Machine Learning Techniques

“…Jha and Parvathi conducted a study that used an SVM classifier trained by extracting features using the histogram of oriented gradient (HOG) feature extraction method to translate handwritten Odia and Hindi text into Braille. Converting Odia [36] and Hindi [25] using this classification technique, 99% and 94.5% accuracies were achieved, respectively, into Braille. Using the same technique, 99% and 80% accuracies were achieved for converting handwritten English and Sinhala documents into Braille text [35].…”

“…with accuracies of (99.95%, 99.95%, 99.85%, 99.85%, 99.80%, 99.64%, and 99.64%), with AUC (0.9899, 0.9997, 0.9992, 0.9495, 0.9615, 0.9495, and 0.9683, respectively) with TPR >90% and TNR >99%, as shown in Figure 4(b). For category-3 (27)(28)(29)(30)(31)(32)(33)(34)(35)(36)(37)(38)(39), Qaaf-Bri Yay ( ‫ے‬ ‫ﻕ‬ ), maximum TA of 100% is achieved for class 35 (Gol Hy), 37 (Hamza), and 28 (Kaaf ) ( ‫ﻩ‬ , ‫ء‬ , ‫ﻙ‬…”

Characterization of English Braille Patterns Using Automated Tools and RICA Based Feature Extraction Methods

“…Braille has been converted into other languages like Urdu [40,41], Arabic [42][43][44], Hindi [19,45], Bengali [46,47] Tamil [22], Sinhala [23], Kannada [48], ODIA [49], Chinese [50,51], Korean [52] and Gujarati [53][54][55].…”

“…This study used robust supervised machine learning techniques like SVM, KNN and Decision Trees to predict English Braille characters correctly. Even with small datasets, these algorithms are well-known for better text prediction [49]. These machine learning techniques were combined with the RICA-based feature extraction method, as it helps improve accuracy, reduce the risk of overfitting the model and speeds up the training process.…”

“…In the same way, Li, Xeng and Zu used conventional extraction methods to identify Braille characters using KNN and the Naïve Bayes classifier [58]. Additionally, Li, Yan and Zhang employed HOG extractor methods using SVM to translate Braille to Odia, Sinhala and English [23,49]. The current study uses the reconstruction Independent Component Analysis (RICA)-and PCA-based feature extraction techniques with SVM, DT and KNN for Braille to English alphabet text translation.…”