Speech Recognition System of Arabic Alphabet Based on a Telephony Arabic Corpus

Abstract

Abstract. Automatic recognition of spoken alphabets is one of the difficult tasks in the field of computer speech recognition. In this research, spoken Arabic alphabets are investigated from the speech recognition problem point of view. The system is designed to recognize spelling of an isolated word. The Hidden Markov Model Toolkit (HTK) is used to implement the isolated word recognizer with phoneme based HMM models. In the training and testing phase of this system, isolated alphabets data sets are taken from the telephony Arabic speech corpus, SAAVB. This standard corpus was developed by KACST and it is classified as a noisy speech database. A hidden Markov model based speech recognition system was designed and tested with automatic Arabic alphabets recognition. Four different experiments were conducted on these subsets, the first three trained and tested by using each individual subset, the fourth one conducted on these three subsets collectively. The recognition system achieved 64.06% overall correct alphabets recognition using mixed training and testing subsets collectively.

Convolutional Neural Network Based Real Time Arabic Speech Recognition to Arabic Braille for Hearing and Visually Impaired

“…Speech to Text is the process of converting or deciphering an audible signal obtained by a receiver or a phone into a string of words. Extraction of highlights from a gathered signal and matching them to a voice test in an information base is used to perceive a word ( 31 , 32 ).…”

Concatenative based Pashto Digits and Numbers Synthesizer

“…In Pashto language the number system is similar to the most further foreign languages for example Persian [2], Arabic [11] and other languages [9], in which every digit from zero to ten has different name, similarly in the Pashto language case every digits from Sefer (0) to Las (10) has different pronunciation. The numbers subsequent to Las (10), Yawo-las (1) to Naha (9) with small variation, whereas from Yaw-Vesht (21) to Naha-Vesht (29) the similar postfix is Vesht with number from Yaw (1) to Naha (9) as a prefix.…”

“…In concatenative synthesis we concatenate the isolated Pashto digits from the speech database, which have been created from a few isolated Pashto digits that contains Sefer (0) to naha (9) then Las (10), Shul (20), Dairsh (30), Celwaikht (40), Panzoos (50), Shpeta (60), Awaya (70), Atya (80), Nawi (90), Sul (100), and Zar (1000). Through concatenation of these few isolated Pashto digits we produce Pashto digits from Yaw-las (11) to Yaw-zar (1000). Figure 3 shows the plot of the signal for the Pashto digit Celour (four), and Figure 4 shows the plot of Dairsh (thirty).…”

“…Celour (four) and Dairsh (thirty) are present in the speech database but Celour Dairsh (thirth four) is not present in the speech database. By concatenating Celour (4) and Dairsh (thirty), Celour Dairsh (34) is produce and similarly other digits too from Yaw Las (11) upto Yaw Zar (1000). For comparing the plot of synthesized concatenative Pashto digits Celour Dairsh is shown in Figure 5 and the plot of recorded Celour Dairsh (34) shown in Figure 6.…”