Overview of Automatic Speech Recognition, Approaches and Challenges: Way the Future to Turkish Speech Recognition

Speech is a paramount means of communication among humans, which makes recognition of the speech by computers is a study area of significance. In this research area, many studies have been carried out based on different languages. Studies in the literature have played an important role in increasing the performance of speech recognition technologies. This study presents a literature review on speech recognition and then discusses the recorded signs of progress made in this area of research for different languages. The data sets used in speech recognition systems, feature extraction approaches, speech recognition methods and performance evaluation criteria are examined and the focus is on the development of speech recognition and the difficulties in this field. Recent studies in the field of speech recognition have focused on developing more powerful methods against adverse conditions (environmental noise, speaker and language variability). Therefore, a general perspective on recent advances in speech recognition in unfavorable conditions is presented. In this way, an attempt has been made to help in selecting the methods that could be used to overcome the congestion and difficulties in speech recognition under adverse conditions. In addition, the well-known methods used in Turkish speech recognition were compared. Difficulty in the recognition of Turkish speech and appropriate methods to overcome these difficulties have been investigated in detail. Accordingly, a future route assessment of Turkish speech recognition has been put forward.