Scene texts which are found in scene images contain valuable meanings. Scene text recognitionrnis the process of converting text regions on the image into machine readable and editable symbols.rnNaturally, scene texts can appear in regular or irregular layout. In scene texts irregular text isrnwidely used. Scene text with an irregular layout is difficult to recognize because of different formsrnof distortions. Correcting these distortions without losing any desired information is one of thernmajor challenges in computer vision. Different approaches are proposed to solve the problem ofrndistortion in scene text recognition. Based on their proposed techniques, these approaches can berncategorized into four categories: Character Level Strong Supervision, Rectification, MultirnDirection Encoding and Attention based approaches. The state of the art is the attention-basedrnapproach which predicts characters from scene text image features by using encoder and decoderrnwith attention methods. The performance of attention-based approaches, however, is not goodrnmainly because they are unable to extract detail image features. The approach underperformsrnparticularly with long scene text sequences due to their inconsistent and decreasing encoder outputrnutilization during each decoding time step. Also, it faces the problem of attention mismatch forrnseverely distorted texts. rn To tackle the problem in attention-based encoder decoder approach, we proposed globalrnattention based mechanism with Bi-LSTM decoder which can handle any type of text distortionsrnimplicitly. The proposed approach is trained with 6,000 regular and irregular scene text imagesrnrandomly taken from publicly available SYN90K synthetic datasets. The dataset is widely used torntrain scene text recognizers. Preprocessing tasks which are image rescaling and noise removal arernperformed only for training purpose. rn The proposed approach is evaluated using 4 class of regular scene text image datasets and 3rnclass of irregular scene text image datasets. The proposed approach outperforms the state-of-theartrnapproachrnbyrnanrnaveragernofrn1.58%rnonrnregularrnscenerntextrnimagerndatasetsrnandrnbyrnanrnaveragernofrnrn1.85%rnonrnirregularrnscenerntextrnimagerndatasets.rnInrnaddition,rnthernincorporationrnofrnBi-LSTMrndecoderrnrninrnthernproposedrnapproachrnincreasesrnthernrecognitionrnperformancernbyrnanrnaveragernofrn5.24%rnforrnregularrnrnscenerntextsrnandrnbyrnanrnaveragernof 3.05%rnforrnirregularrnscenerntexts.