policy

KON-Artist

MLsound · PyTorch

or hover any field below to flag it

Overview

Name
KON-Artist
Author
MLsound
Framework
PyTorch
License
unknown
Skill type
other
Evidence level
untested
Task description
A RL framework designed to bypass the AASIST3 spoofing detector. It optimizes audio generators to exploit the spline-based vulnerabilities of KAN architectures. By treating AASIST3 as a frozen reward signal, the agent learns to hide synthetic artifacts within the model's mathematical blind spot.

Spaces

Action space
other · 0-dim · 0Hz
Observation space
  • type: other

Links

HuggingFace repo
null
Paper (arXiv)
null

Compatible robots

3+17 mentioned but not in catalog yet

Compatible environments

0

No environments list KON-Artist yet.

Datasets that reference this policy

0

No datasets reference KON-Artist yet.