AI RESEARCH

Learn2Synth: Learning Optimal Data Synthesis Using Hypergradients for Brain Image Segmentation

arXiv CS.LG

ArXi:2411.16719v4 Announce Type: replace-cross Domain randomization through synthesis is a powerful strategy to train networks that are unbiased with respect to the domain of the input images. Randomization allows networks to see a virtually infinite range of intensities and artifacts during