← Search

Yaël Balbastre

2 accepted papers

2025

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

ICCV 2025poster

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 training, thereby minimizing overfitting to appearance…

2025

Schwarz–Schur Involution: Lightspeed Differentiable Sparse Linear Solvers

ICML 2025poster

Sparse linear solvers are fundamental to science and engineering, applied in partial differential equations (PDEs), scientific computing, computer vision, and beyond. Indirect solvers possess characteristics that make them undesirable as stable differentiable modules; existing direct solvers, though…

Cited by 0SourcePDFScholar