Full Professor, University of Michigan - Ann Arbor
2 papers at NeurIPS 2025
Deep learning image filters lack normalization. We propose filter normalization with learnable scaling and shifting, achieving significant improvements in intensity variation benchmarks.
