• Publications
  • Influence
Turning Your Weakness Into a Strength: Watermarking Deep Neural Networks by Backdooring
TLDR
We present an approach for watermarking Deep Neural Networks in a black-box way. Expand
  • 104
  • 34
  • PDF
Fine-grained Analysis of Sentence Embeddings Using Auxiliary Prediction Tasks
TLDR
We propose a framework that facilitates fine-grained measurement of some of the information encoded in sentence embeddings. Expand
  • 259
  • 29
  • PDF
Houdini: Fooling Deep Structured Prediction Models
TLDR
We introduce a novel flexible approach named Houdini for generating adversarial examples specifically tailored for the final performance measure of the task considered, be it combinatorial and non-decomposable. Expand
  • 142
  • 14
  • PDF
Fooling End-To-End Speaker Verification With Adversarial Examples
TLDR
Adversarial examples are generated by adding a peculiar noise to original speaker examples, in such a way that they are almost indistinguishable, by a human listener. Expand
  • 68
  • 3
  • PDF
Out-of-Distribution Detection using Multiple Semantic Label Representations
TLDR
We propose to use multiple semantic dense representations instead of sparse representation as the target label. Expand
  • 31
  • 2
  • PDF
Voice Separation with an Unknown Number of Multiple Speakers
TLDR
We present a new method for separating a mixed audio sequence, in which multiple voices speak simultaneously, while maintaining the speaker in each output channel fixed. Expand
  • 26
  • 2
  • PDF
Houdini: Fooling Deep Structured Visual and Speech Recognition Models with Adversarial Examples
TLDR
We introduce a novel flexible approach named Houdini for generating adversarial examples specifically tailored for the final performance measure of the task considered, be it combinatorial and non-decomposable. Expand
  • 32
  • 1
  • PDF
The Influence of Lexical Selection Disruptions on Articulation
Interactive models of language production predict that it should be possible to observe long-distance interactions; effects that arise at one level of processing influence multiple subsequent stagesExpand
  • 7
  • 1
  • PDF
Hide and Speak: Deep Neural Networks for Speech Steganography
TLDR
In this paper, we explore the use of deep neural networks as steganographic functions for speech data. Expand
  • 3
  • 1
Unsupervised Cross-Domain Singing Voice Conversion
TLDR
We present a wav-to-wav generative model for the task of singing voice conversion from any identity. Expand
  • 3
  • 1
  • PDF