linkding

#attack_vectors #audio #neural_networks #security #speech_recognition | <blockquote>We have constructed targeted audio adversarial examples on speech-to-text transcription neural networks: given an arbitrary waveform, we can make a small perturbation that when added to the original waveform causes it to transcribe as any phrase we choose.</blockquote>

7 years ago | View Shared by victor

LINKDING

Shared bookmarks

User

Tags