Robust Deep Multi-Modal Sensor Fusion using Fusion Weight Regularization and Target Learning
This work addresses robustness issues in sensor fusion for applications like health care and autonomous systems, representing an incremental improvement over existing deep learning techniques.
The paper tackles the problem of inconsistent fusion weights in deep multi-modal sensor fusion, particularly under sensor failures, by proposing regularized gating architectures with fusion weight regularization and target learning, resulting in outperformance over existing methods, especially when sensory inputs are corrupted.
Sensor fusion has wide applications in many domains including health care and autonomous systems. While the advent of deep learning has enabled promising multi-modal fusion of high-level features and end-to-end sensor fusion solutions, existing deep learning based sensor fusion techniques including deep gating architectures are not always resilient, leading to the issue of fusion weight inconsistency. We propose deep multi-modal sensor fusion architectures with enhanced robustness particularly under the presence of sensor failures. At the core of our gating architectures are fusion weight regularization and fusion target learning operating on auxiliary unimodal sensing networks appended to the main fusion model. The proposed regularized gating architectures outperform the existing deep learning architectures with and without gating under both clean and corrupted sensory inputs resulted from sensor failures. The demonstrated improvements are particularly pronounced when one or more multiple sensory modalities are corrupted.