Search papers, labs, and topics across Lattice.
Tallinn University of Technology, Department of Computer systems
2
1
3
10
Pruning neural networks for homomorphic encryption can be done safely, reducing catastrophic accuracy drops by up to 29 times while enhancing efficiency.
Quantizing neural networks doesn't have to mean sacrificing robustness: a new three-stage framework achieves up to 10.35% better attack resilience and 12.47% better fault resilience.
