Crying is one of the most fundamental ways an infant can communicate with the outside world. The cry contains vital information to determine the needs of the baby, whether due to hunger, pain, fatigue, or simply discomfort [1]. Automated infant cry classification is crucial for early diagnosis and proper care. The existing methods are constrained by two factors: Inability to adapt to cry variations and lack of clinical transparency. This study presents an innovative approach using explainable reinforcement learning and feature fusion methods which dynamically assigns different attention weights to already extracted features using a lightweight policy agent learnt via the REINFORCE algorithm [2]. The model is trained and validated on a widely used literature dataset named Donate-a-Cry Corpus, which classifies cries into five categories namely; hunger, tiredness, belly pain, burping, and discomfort. This paper also introduces a dynamic reward shaping mechanism into the reinforcement loop that improves the agent’s ability to focus on underrepresented classes. To validate our model, k=3-fold cross-validation is applied to achieve an accuracy of 94.44%.
