It is a very difficult task to manually process the handwritten documents due to varieties of handwritten scripts and lack of associated language dictionary to interpret documents. Most of the large companies as well as small-scale industries want to automate the process of script recognition. The big challenge is to make machines recognize the hand-printed scripts. Humans can recognize handwritten or hand-printed words after gaining knowledge of a specific language. In the same way, machines should be trained to recognize the handwritten scripts. This process of transferring human knowledge to computers should be automated. The proposed research work attempts to automate the character recognition system for Devanagari script using various machine learning classifiers like Decision Tree classifier, Nearest Centroid classifier, K Nearest Neighbors classifier, Extra Trees classifiers and Random Forest classifier. The performance of all the classifiers is evaluated using accuracy parameter as success criteria. The Extra Trees classifiers and Random Forest classifier is proved to better than other classifiers with 78% and 77% of accuracy respectively. The robustness to picture quality, writing style, font size is the novelty of the OCR system which makes it ideal to use.

Character Recognition System for Devanagari Script Using Machine Learning Approach

Systems that extract information from speech are common in communications and automation, where an individual's voice is analysed and recognised by the system to follow the intended course of action. Extant research on speech recognition deals with the application of complex performance techniques of Artificial Intelligence (AI) and Machine Learning (ML) to study large non-isolated voice patterns. This paper uses simplistic performance parameters, namely, cross correlation for the identification of individual voice samples consisting of isolated Marathi and Konkani words in MATLAB. The study examines a dataset of 90 samples comprising 10 words from both languages for 10 individuals. The study presents a comparative analysis of. wav and. ogg formats of the audio samples with efficiencies of 93.5% and 88.5% for Marathi and Konkani respectively.

Marathi and Konkani Speech Recognition using Cross-Correlation Analysis

This paper demonstrates a system for transforming the text inscribed in Konkani dialect into Speech by utilizing artificial neural network. Many visually challenged individuals use Text to Speech framework as a tool for communication. The ability to convert text to voice lessens the reliance, dissatisfaction, and feeling of defenselessness of these individuals. India is called as the land of unity and diversity, there are 22 official languages. TTS frameworks are mostly accessible in English; in any case, it has been watched that individuals feel more comfortable in hearing their own native dialect. Handwritten optical character recognition is the most challenging research zone, because of its intricacy in segmenting the character that grows on account of Devnagari Script because of Modifiers and compound characters. The document comprising Konkani text is scanned and fed to the system. In this framework the character recognition is done by utilizing Neural Network, in this manner the structure can be upgraded to work with letters written in different styles. After the characters in the Documents are viably recognized by neural network, it is composed to a text document, the entered text document is analyzed, the syllabification is accomplished in view of the phonological guidelines and the syllables are secured autonomously. At that point the syllable coordinating speech file is linked and the silence existing in the linked discourse is confined. breaks within the discourse are removed at syllable limits without diminishing the superiority of speech.

Ramita Pradeep Karpe

Papers

Konkani Script to Speech Conversion by Concatenation of recognized Hand written Konkani Text Using Neural Network