Research map / Architectures and efficiency / CNN design
Foveated BagNet: Inherent Interpretability Does Not Exclude Global Context
Holger Heidrich, Sarah Müller, Andreas Schilling
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes FovBagNet, a foveated extension of BagNet that adds global context to interpretable-by-design patch-based classifiers while preserving inherent explainability.
Problem
Interpretable-by-design models like BagNet restrict each local class predictor to small image patches, yielding explanations but incurring a substantial accuracy penalty because large-scale image features cannot be captured.
Approach
Introduces a foveated architecture on top of BagNet that incorporates global context while keeping predictions decomposable into local patch-level contributions, so interpretability is retained without sacrificing large-scale feature capture.
Experimental setup
Baselines: BagNet
Builds on: BagNet
Closest papers by meaning
- Approximating CNNs with Bag-of-local-Features models works surprisingly well on ImageNet (ICLR 2019 · similarity 0.88)
- TransNeXt: Robust Foveal Visual Perception for Vision Transformers (CVPR 2024 · similarity 0.85)
- Seeing More with Less: Human-like Representations in Vision Models (CVPR 2025 · similarity 0.85)
- DD-CAM: Minimal Sufficient Explanations for Vision Models Using Delta Debugging (NeurIPS 2026 · similarity 0.84)
- Explicitly Modeling Subcortical Vision with a Neuro-Inspired Front-End Improves CNN Robustness (NeurIPS 2025 · similarity 0.84)
- CanViT: Toward Active-Vision Foundation Models (NeurIPS 2026 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
