Hierarchical Invariance for Robust and Interpretable Vision Tasks at Larger Scales
This addresses the need for robust and interpretable vision systems for trustworthy AI applications, though it appears to be an incremental improvement combining existing concepts like invariance and hierarchical structures.
The paper tackles the problem of limited discriminability in invariant representations for vision tasks by proposing hierarchical invariance, which constructs over-complete invariants with a CNN-like architecture while maintaining interpretability. It demonstrates competitive accuracy, invariance, and efficiency in texture, digit, and parasite classification, and shows effectiveness in real-world applications like adversarial perturbation detection and AIGC forensics.
Developing robust and interpretable vision systems is a crucial step towards trustworthy artificial intelligence. In this regard, a promising paradigm considers embedding task-required invariant structures, e.g., geometric invariance, in the fundamental image representation. However, such invariant representations typically exhibit limited discriminability, limiting their applications in larger-scale trustworthy vision tasks. For this open problem, we conduct a systematic investigation of hierarchical invariance, exploring this topic from theoretical, practical, and application perspectives. At the theoretical level, we show how to construct over-complete invariants with a Convolutional Neural Networks (CNN)-like hierarchical architecture yet in a fully interpretable manner. The general blueprint, specific definitions, invariant properties, and numerical implementations are provided. At the practical level, we discuss how to customize this theoretical framework into a given task. With the over-completeness, discriminative features w.r.t. the task can be adaptively formed in a Neural Architecture Search (NAS)-like manner. We demonstrate the above arguments with accuracy, invariance, and efficiency results on texture, digit, and parasite classification experiments. Furthermore, at the application level, our representations are explored in real-world forensics tasks on adversarial perturbations and Artificial Intelligence Generated Content (AIGC). Such applications reveal that the proposed strategy not only realizes the theoretically promised invariance, but also exhibits competitive discriminability even in the era of deep learning. For robust and interpretable vision tasks at larger scales, hierarchical invariant representation can be considered as an effective alternative to traditional CNN and invariants.