lcnn-v7-cqt

lcnn-v7-cqt — Improved LCNN for ASVspoof 2019 LA

Best improved LCNN checkpoint from a comparative study of three neural anti-spoofing architectures. Supersedes caa-speech-detection-asvspoof2019/lcnn.

Version: lcnn_v7 (CQT + label-smoothing + cosine schedule + grad-clip)

Architecture

Light CNN over Constant-Q Transform (CQT) spectrograms.

Reference: Lavrentyeva et al., "Audio Replay Attack Detection with Deep Learning Frameworks", Interspeech 2017.

Training

Dataset: ASVspoof 2019 LA train split (~25k utterances). No data augmentation.

Results

Baseline to beat: EER 8.09% (LFCC+GMM).

Dev EER improved from 0.902% (baseline LCNN) to 0.708%. Eval EER improved from 8.43% to 3.26% — a 61% relative reduction.

See learning_curves/lcnn_v4_cqt_vs_v7.png for the training trajectory comparing v4_cqt and v7.

Usage

Install dependencies from the source repository, then:

import torch
from src.models.lcnn.model import LCNNModel

config = {
    "sample_rate": 16000,
    "transform_type": "cqt",
    "n_bins": 84,
    "bins_per_octave": 12,
    "hop_length": 160,
    "fmin": 32.7,
    "max_audio_seconds": 4.0,
    "channels": [32, 48, 64, 128],
    "kernel_sizes": [5, 5, 3, 3],
    "fc_hidden": 64,
    "dropout": 0.3,
    "label_smoothing": 0.1,
}

model = LCNNModel(config)
state = torch.load("best.pt", map_location="cpu")
model.load_state_dict(state)
model.eval()

# waveform: (B, T) float32 at 16 kHz, T up to 64000
with torch.no_grad():
    logits = model({"waveform": waveform})["logits"]
    probs = torch.softmax(logits, dim=-1)  # [:, 0] = spoof, [:, 1] = bonafide

Source: github.com/sebastiaoteixeira/caa-ai-generated-speech-detector

Limitations

• Trained on ASVspoof 2019 LA only. Not evaluated on physical access or other corpora.
• In-the-Wild EER of 33.41% indicates limited generalisation to real-world spoofing not captured in ASVspoof 2019.
• No data augmentation was applied; domain shift to real deployments may degrade performance.

Citation

@inproceedings{wang2020asvspoof,
  title     = {{ASVspoof} 2019: A Large-Scale Public Database of Synthesized, Converted and Replayed Speech},
  author    = {Wang, Xin and others},
  booktitle = {Computer Speech \& Language},
  volume    = {64},
  year      = {2020}
}

@inproceedings{lavrentyeva2017audio,
  title     = {Audio Replay Attack Detection with Deep Learning Frameworks},
  author    = {Lavrentyeva, Galina and Novoselov, Sergey and Malykh, Egor and Kozlov, Alexander and Kudashev, Oleg and Shchemelinin, Vadim},
  booktitle = {Interspeech},
  year      = {2017}
}