Postdoc, University of California, Berkeley
1 paper 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.
