{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.10.13","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"none","dataSources":[{"sourceId":73047,"databundleVersionId":8149390,"sourceType":"competition"}],"dockerImageVersionId":30684,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":false}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"## Competition Citation\n\n@misc{ben10,\n    author = {Md. Rezuwan Hassan, Mohaymen Ul Anam, Rubayet Sabbir Faruque, S M Jishanul Islam, Sushmit, Tahsin},\n    title = {ভাষা-বিচিত্রা: ASR for Regional Dialects},\n    publisher = {Kaggle},\n    year = {2024},\n    url = {https://kaggle.com/competitions/ben10}\n}","metadata":{}},{"cell_type":"markdown","source":"# Introduction\n\nWelcome to our notebook dedicated to tackling the Bengali speech recognition challenge hosted by Bengali.AI. In this competition, our goal is to develop a robust solution system capable of accurately transcribing Bengali speech across various regional dialects, adhering to linguistically accurate orthography. This competition presents a unique opportunity to contribute to the advancement of Bengali speech recognition technology, particularly in handling the rich diversity of dialects prevalent in Bangladesh.\n\nThe provided dataset, curated by Bengali.AI, comprises recordings from 373 individuals spanning ten different geographical regions, including Rangpur, Kishoreganj, Narail, Chittagong, Narsingdi, Tangail, Barishal, Habiganj, Sylhet, and Sandwip. These recordings capture spontaneous speech on everyday topics, totaling over 79 hours of audio data. Leveraging this extensive corpus, participants are tasked with pioneering open-source speech recognition methods for Bengali, aiming to bridge the gap between modern speech recognition systems and the linguistic intricacies of local dialects.\n\nThroughout this notebook, we will embark on a journey to explore the dataset, preprocess the audio and text data, design and train a suitable speech recognition model, and evaluate its performance. Our ultimate aim is to develop a model with high accuracy in transcribing Bengali speech, thus contributing to the advancement of speech recognition technology for the Bengali language. ","metadata":{}},{"cell_type":"markdown","source":"# Importing Libraries","metadata":{}},{"cell_type":"code","source":"import pandas as pd\nimport matplotlib.pyplot as plt","metadata":{"execution":{"iopub.status.busy":"2024-04-10T10:59:33.079609Z","iopub.execute_input":"2024-04-10T10:59:33.080351Z","iopub.status.idle":"2024-04-10T10:59:34.301907Z","shell.execute_reply.started":"2024-04-10T10:59:33.080308Z","shell.execute_reply":"2024-04-10T10:59:34.300655Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Data Exploration and Preprocessing","metadata":{}},{"cell_type":"code","source":"train_csv = pd.read_csv(\"/kaggle/input/ben10/ben10/train.csv\")\ntrain_csv.head()","metadata":{"execution":{"iopub.status.busy":"2024-04-10T10:59:34.303838Z","iopub.execute_input":"2024-04-10T10:59:34.304251Z","iopub.status.idle":"2024-04-10T10:59:34.585739Z","shell.execute_reply.started":"2024-04-10T10:59:34.304223Z","shell.execute_reply":"2024-04-10T10:59:34.584593Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Model Building","metadata":{}},{"cell_type":"markdown","source":"# Model Training","metadata":{}},{"cell_type":"markdown","source":"# Model Evaluation","metadata":{}},{"cell_type":"markdown","source":"# Model Testing and Submission","metadata":{}},{"cell_type":"markdown","source":"# Conclusion","metadata":{}}]}