Masked Autoencoders that Listen

Open paper
Benchmark Model Rank Results
speaker-identification-on-voxceleb1AudioMAE (local)#4Top-1 (%): 94.8Accuracy: 94.8
speaker-identification-on-voxceleb1AudioMAE (global)#7Top-1 (%): 94.1Accuracy: 94.1

Rank counts only results with a code link.