Towards Robust Transcription: Exploring Noise Injection Strategies for Training Data Augmentation

For instance, when employing noise injection, several key factors must be considered, Recent advancements in Automatic Piano Transcription such as the type of noise (e.g., white, pink, environmental), (APT) have significantly improved system performance, the Signal-to-Noise-Ratio (SNR), and the ratio of clean to but the impact of noisy environments on the system performance augmented data. However, to the best of our knowledge, remains largely unexplored. This study investigates these parameters are often chosen arbitrarily, highlighting the impact of white noise at various Signal-to-Noise Ratio the need for further investigation in this area.