Olfactory Prediction Using CNN-CBAM Model

Olfactory perception prediction plays a vital role in multi-modal sensory research, offering insights for health monitoring and personalized experiences. In this work, we propose a novel CNN-CBAM model that combines Convolutional Neural Network (CNN) with a Channel-Spatial Attention Mechanism (CBAM) to enhance olfactory prediction accuracy. The CNN extracts deep features from multi-modal data, while CBAM improves the model's focus on relevant information through adaptive attention across channels and spatial dimensions. Experimental results show that the CNN-CBAM model significantly outperforms traditional models like ConvLSTM and CNN, demonstrating superior prediction performance with improved robustness.