The task of automatically recognizing musical instruments poses significant challenges within the domain of music information retrieval. Learning to play the piano, on the other hand, demands expert instruction and substantial practice. Due to the hectic nature of modern life, many individuals find it difficult to commit to systematic training. Additionally, the scarcity of qualified piano teachers and the high costs associated with lessons further discourage potential students. If a computer could recognize and assess a learner's piano performance in real time, it would enable learners to identify and correct their mistakes promptly. Although there are existing music recognition technologies, most suffer from several limitations. Currently, music processing systems that incorporate models for chord progressions achieve high accuracy in tasks such as music structure analysis, multi pitch analysis, and automatic composition or accompaniment. pitch patterns are treated as observations derived from the hidden states within the chord progression model. Convolutional Neural Networks (CNN) have been successfully applied to chord recognition. The CNN approch will give high accuracy, precision and F1-Score.