Layer-Wise Modality Decomposition for Interpretable Multimodal Sensor Fusion
This addresses the need for transparency in autonomous driving perception models to prevent catastrophic misperceptions, though it is incremental as it builds on existing interpretability methods for a specific domain.
The paper tackled the problem of interpreting multimodal sensor fusion models in autonomous driving by introducing Layer-Wise Modality Decomposition (LMD), a post-hoc, model-agnostic method that disentangles modality-specific contributions across layers, validated on camera-radar, camera-LiDAR, and camera-radar-LiDAR settings with structured perturbation-based metrics and visual decompositions.
In autonomous driving, transparency in the decision-making of perception models is critical, as even a single misperception can be catastrophic. Yet with multi-sensor inputs, it is difficult to determine how each modality contributes to a prediction because sensor information becomes entangled within the fusion network. We introduce Layer-Wise Modality Decomposition (LMD), a post-hoc, model-agnostic interpretability method that disentangles modality-specific information across all layers of a pretrained fusion model. To our knowledge, LMD is the first approach to attribute the predictions of a perception model to individual input modalities in a sensor-fusion system for autonomous driving. We evaluate LMD on pretrained fusion models under camera-radar, camera-LiDAR, and camera-radar-LiDAR settings for autonomous driving. Its effectiveness is validated using structured perturbation-based metrics and modality-wise visual decompositions, demonstrating practical applicability to interpreting high-capacity multimodal architectures. Code is available at https://github.com/detxter-jvb/Layer-Wise-Modality-Decomposition.