APEX: Audio Prototype EXplanations for Classification Tasks
For researchers and practitioners needing interpretable audio AI, APEX addresses the lack of audio-specific XAI methods by offering a prototype-based approach that respects acoustic multidimensionality without requiring model fine-tuning.
APEX introduces a post-hoc, training-free framework for explaining pre-trained audio classifiers by disentangling explanations into four acoustic perspectives (square, time, frequency, time-frequency), preserving output invariance and providing more intuitive, example-based explanations than gradient-based methods.
Explainable AI (XAI) has achieved remarkable success in image classification, yet the audio domain lacks equally mature solutions. Current methods apply vision-based attribution techniques to spectrograms, overlooking fundamental differences between visual and acoustic signals. While prototype reasoning is promising, acoustic similarity remains multidimensional. We introduce APEX (Audio Prototype EXplanations), a post-hoc framework for interpreting pre-trained audio classifiers. Crucially, APEX requires no fine-tuning of the original backbone and strictly preserves output invariance. APEX disentangles explanations into four perspectives: Square-based prototypes to localize transient events, Time-based for temporal patterns, Frequency-based highlighting spectral bands, and Time-Frequency-based integrating both. This yields intuitive, example-based explanations that respect acoustic properties, providing greater semantic clarity than standard gradient-based methods.