Visual Concept Connectome (VCC): Open World Concept Discovery and their Interlayer Connections in Deep Models
This addresses the fundamental challenge of interpretability in computer vision for researchers and practitioners, though it is incremental as it builds on prior single-layer concept extraction methods.
The paper tackles the problem of understanding what deep network models capture in their learned representations by introducing the Visual Concept Connectome (VCC), a methodology that discovers human-interpretable concepts and their interlayer connections in an unsupervised manner, with quantitative and qualitative results showing its effectiveness in image classification.
Understanding what deep network models capture in their learned representations is a fundamental challenge in computer vision. We present a new methodology to understanding such vision models, the Visual Concept Connectome (VCC), which discovers human interpretable concepts and their interlayer connections in a fully unsupervised manner. Our approach simultaneously reveals fine-grained concepts at a layer, connection weightings across all layers and is amendable to global analysis of network structure (e.g., branching pattern of hierarchical concept assemblies). Previous work yielded ways to extract interpretable concepts from single layers and examine their impact on classification, but did not afford multilayer concept analysis across an entire network architecture. Quantitative and qualitative empirical results show the effectiveness of VCCs in the domain of image classification. Also, we leverage VCCs for the application of failure mode debugging to reveal where mistakes arise in deep networks.