Variable-Radius Disk Transforms and an Area-Integral Problem of Zalcman
arXiv:2608.02546
2026
Architecture
1 ideas extracted · analyzed Aug 31, 2026
What the math gives to ML
The paper studies a spatially nonstationary disk-integral transform whose receptive-field radius shrinks linearly with distance to the boundary. Its key transferable property is injectivity: in the continuum, recording the local disk integral at every center does not necessarily destroy information, despite the operation being a smoothing transform. This suggests a boundary-aware convolution or pooling layer for image and feature maps, paired with reconstruction to prevent the discretized transform from collapsing high-frequency information. The theorem is exact only in specified continuous function spaces, so finite-grid implementations should measure conditioning and reconstruction error explicitly.
Ideas from this paper
Unverified
2026
Replace fixed-radius image blur or pooling with disk averages whose radius is proportional to the distance from each pixel to the image boundary. Compute the transform at every spatial location and train a lightweight decoder to reconstruct the pre-transform feature map, using reconstruction error as an anti-collapse regularizer. This creates a scale-adaptive smoothing layer with an injectivity motivation in the continuum while providing larger context in the image interior.
Useful5/10
Difficulty5/10
Novelty7/10