List your company

Automatic Speech Recognition on LibriSpeech (test)

1.97WER (%)Clean
1.3085.776510.24514.7135Jun 4, 2026Jun 5, 2026

Evaluation results

MethodLinks
CleanPrefix (s)=4.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.061.97——
CleanPrefix (s)=4.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.061.98——
CleanPrefix (s)=2.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.062.01——
CleanPrefix (s)=3.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.062.01——
CleanPrefix (s)=4.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.062.01——
CleanPrefix (s)=1.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.062.02——
CleanPrefix (s)=2.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.062.02——
CleanPrefix (s)=3.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.062.02——
CleanPrefix (s)=2.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.062.03——
CleanPrefix (s)=3.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.062.03——
CleanPrefix (s)=1.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.062.05——
CleanPrefix (s)=1.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.062.05——
AO*Prefix (s)=2.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.062.16——
AO*Prefix (s)=3.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.062.35——
AO*Prefix (s)=2.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.063.19——
AO*Prefix (s)=3.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.064.27——
AO*Prefix (s)=2.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.064.38——
Clean (No Attack)Defense=None (Raw Model)2026.064.75——
Clean (No Attack)Defense=PVP Vote, Defense Type=Input Preprocessing2026.064.8——
AO*Prefix (s)=4.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.064.84——
AO*Prefix (s)=1.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.065.34——
Clean (No Attack)Defense=DT-AT, Defense Type=Adversarial Training2026.065.7——
Clean (No Attack)Defense=CE-AT, Defense Type=Adversarial Training2026.065.72——
AO*Prefix (s)=4.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.065.8——
Clean (No Attack)Defense=TI, Defense Type=Adversarial Training2026.065.81——
Clean (No Attack)Defense=DMW, Defense Type=Adversarial Training2026.065.86——
Muting WhisperDefense=DMW, Defense Type=Adversarial Training2026.065.94——
Muting WhisperDefense=TI, Defense Type=Adversarial Training2026.066.04——
Muting WhisperDefense=DT-AT, Defense Type=Adversarial Training2026.066.09——
Muting WhisperDefense=CE-AT, Defense Type=Adversarial Training2026.066.11——
Clean (No Attack)Defense=AudioPure, Defense Type=Input Preprocessing2026.066.17——
Clean (No Attack)Defense=LPF, Defense Type=Input Preprocessing2026.066.45——
Clean (No Attack)Defense=WaveGuard, Defense Type=Input Preprocessing2026.066.76——
Muting WhisperDefense=AudioPure, Defense Type=Input Preprocessing2026.066.8——
Muting WhisperDefense=LPF, Defense Type=Input Preprocessing2026.067.52——
Muting WhisperDefense=WaveGuard, Defense Type=Input Preprocessing2026.067.62——
PGDDefense=DT-AT, Defense Type=Adversarial Training2026.068.13——
AO*Prefix (s)=1.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.068.28——
PGDDefense=AudioPure, Defense Type=Input Preprocessing2026.068.38——
AO*Prefix (s)=3.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.068.45——
PGDDefense=CE-AT, Defense Type=Adversarial Training2026.068.69——
SlothDefense=AudioPure, Defense Type=Input Preprocessing2026.068.7——
SlothDefense=DT-AT, Defense Type=Adversarial Training2026.068.71——
AO*Prefix (s)=4.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.068.78——
SlothDefense=DMW, Defense Type=Adversarial Training2026.069.26——
SlothDefense=CE-AT, Defense Type=Adversarial Training2026.069.38——
SlothDefense=TI, Defense Type=Adversarial Training2026.069.38——
PGDDefense=DMW, Defense Type=Adversarial Training2026.069.61——
PGDDefense=TI, Defense Type=Adversarial Training2026.069.8——
UniversalPrefix (s)=3.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.069.99——
UniversalPrefix (s)=4.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0610.05——
UniversalPrefix (s)=2.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0610.46——
UniversalPrefix (s)=2.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0610.62——
AOPrefix (s)=3.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0611.11——
UniversalPrefix (s)=4.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0611.15——
SGPrefix (s)=3.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0611.32——
PGDPrefix (s)=4.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0611.56——
UniversalPrefix (s)=4.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0611.89——
PGDPrefix (s)=3.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0612.02——
PGDPrefix (s)=4.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0612.13——
PGDPrefix (s)=4.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0612.18——
AO*Prefix (s)=1.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0612.23——
GT (Bound)Prefix (s)=4.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0612.44——
AOPrefix (s)=2.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0612.58——
PGDPrefix (s)=3.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0612.7——
GT (Bound)Prefix (s)=3.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0612.94——
UniversalPrefix (s)=3.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0613.14——
PGDPrefix (s)=2.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0613.17——
PGDPrefix (s)=3.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0613.21——
AOPrefix (s)=2.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0613.42——
PGDPrefix (s)=2.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0613.75——
AOPrefix (s)=2.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0614.17——
GT (Bound)Prefix (s)=4.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0614.27——
UniversalPrefix (s)=2.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0614.45——
PGDPrefix (s)=2.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0614.46——
GT (Bound)Prefix (s)=4.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0614.46——
GT (Bound)Prefix (s)=2.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0614.51——
AOPrefix (s)=1.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0614.52——
PGDPrefix (s)=1.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0614.58——
AOPrefix (s)=3.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0614.82——
SlothDefense=LPF, Defense Type=Input Preprocessing2026.0614.88——
UniversalPrefix (s)=1.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0614.91——
SlothDefense=WaveGuard, Defense Type=Input Preprocessing2026.0614.98——
AOPrefix (s)=4.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0615.03——
AOPrefix (s)=1.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0615.15——
PGDDefense=WaveGuard, Defense Type=Input Preprocessing2026.0615.47——
PGDPrefix (s)=1.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0615.64——
GT (Bound)Prefix (s)=1.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0616——
MI-FGSMDefense=CE-AT, Defense Type=Adversarial Training2026.0616.2——
SGPrefix (s)=1.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0616.26——
PGDDefense=LPF, Defense Type=Input Preprocessing2026.0616.28——
GT (Bound)Prefix (s)=3.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0616.47——
MI-FGSMDefense=TI, Defense Type=Adversarial Training2026.0616.85——
MI-FGSMDefense=DMW, Defense Type=Adversarial Training2026.0616.86——
MI-FGSMDefense=DT-AT, Defense Type=Adversarial Training2026.0616.98——
PGDPrefix (s)=1.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0617.01——
UniversalPrefix (s)=1.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0617.21——
UniversalPrefix (s)=1.0, Delay (s)=0.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0617.37——
AOPrefix (s)=4.0, Delay (s)=1.0, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0617.68——
UniversalPrefix (s)=3.0, Delay (s)=0.5, SNR constraint=20 dB, Target window=3.0 s, Victim model=Wav2Vec 2.02026.0618.52——