Search papers, labs, and topics across Lattice.
1
0
3
2
By intelligently reorienting adversarial style perturbations, SRasP achieves state-of-the-art cross-domain few-shot learning performance, demonstrating the power of stabilized adversarial training for domain generalization.
