Police: Provably Optimal Linear Constraint Enforcement For Deep Neural Networks.

Randall Balestriero, Yann LeCun

Browse the full ICASSP paper archive.