---
license: other
tags:
  - audio
  - anti-spoofing
  - deepfake-detection
  - speech-anti-spoofing-arena
library_name: pytorch
---

# ProSDD-ASV5: Speech Anti-Spoofing Arena wrapper

Packaged for the [Speech Anti-Spoofing Arena](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena) from the authors' [code](https://github.com/ProSDD/codes/tree/bdb80440cd946368e77f4b51f6ddf26c7443dc33) and [paper](https://arxiv.org/abs/2604.13229).

ProSDD (Mahapatra et al., 2026): a fully fine-tuned wav2vec 2.0 XLS-R 300M encoder, first trained (Stage I, bona fide LibriSpeech train-clean-100 + dev only) to predict masked frame-level speaker + prosody targets, then (Stage II) jointly fine-tuned with that objective and a binary real/fake head on ASVspoof 2024 (ASVspoof 5) train/dev (RawBoost series 3). Inference uses only the clean path: XLS-R CNN front-end + feature projection, latent sequence fixed at 200 frames (zero-padded), 24-layer transformer encoder, mean pool, Linear(1024,512)-ReLU-Linear(512,2). Input as the authors' data_utils_eval.py: 16 kHz, no amplitude normalisation, fixed 4 s (64000 samples) centre crop, shorter clips centre zero-padded. FP32; score = logits[:,1] (bona fide class). Authors' released checkpoint; no upstream license, so the weights are linked from the authors' Google Drive, not re-hosted.

- Weights: **not re-hosted** (the authors grant no license for them). Get the authors' [checkpoint](https://drive.google.com/file/d/1AyD4f0RVVyDncqYbhi8QaP6-GFZ_ZGZl/view); its sha256 is in the `meta.yaml` notes.
- Params: 315.963522 M. Trained on LibriSpeech (stage 1) and ASVspoof 5 train/dev (stage 2).
- Score: higher = bona fide. J-SPAW_LA was scored with the PyTorch wrapper; the other datasets with the
  TensorRT drop-in class (FP32 on the RTX 4070 Ti SUPER; FP16 on the RTX 5060 Ti), which matched PyTorch on the full J-SPAW_LA set.

## Arena results

[![arena tier](https://img.shields.io/endpoint?url=https://speechantispoofingbenchmarks-speechantispoofingarena.hf.space/badge/prosdd-asv5/tier.json)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![arena rank](https://img.shields.io/endpoint?url=https://speechantispoofingbenchmarks-speechantispoofingarena.hf.space/badge/prosdd-asv5/rank.json)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)

EER% (lower is better); EmoSpoofTTS and LRLspoof are spoof-only and use 1-SRR% at the model's
DeepVoice EER threshold. Per-dataset scores are under `.eval_results/`.

[![EER% 19.04 on ASVspoof2019_LA](https://img.shields.io/badge/EER%25%20on%20ASVspoof2019__LA-19.04%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 19.55 on ASVspoof2021_LA](https://img.shields.io/badge/EER%25%20on%20ASVspoof2021__LA-19.55%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 15.73 on ASVspoof2021_DF](https://img.shields.io/badge/EER%25%20on%20ASVspoof2021__DF-15.73%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 7.38 on ASVspoof5](https://img.shields.io/badge/EER%25%20on%20ASVspoof5-7.38%25-yellow)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 6.57 on InTheWild](https://img.shields.io/badge/EER%25%20on%20InTheWild-6.57%25-yellow)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 24.4 on ADD22_eval_31](https://img.shields.io/badge/EER%25%20on%20ADD22__eval__31-24.4%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 28.72 on ADD2023_track12_test_r1](https://img.shields.io/badge/EER%25%20on%20ADD2023__track12__test__r1-28.72%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 28.7 on CFAD](https://img.shields.io/badge/EER%25%20on%20CFAD-28.7%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 4.31 on CD-ADD](https://img.shields.io/badge/EER%25%20on%20CD--ADD-4.31%25-green)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 30.46 on CVoiceFake_small](https://img.shields.io/badge/EER%25%20on%20CVoiceFake__small-30.46%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 10.24 on DeepVoice](https://img.shields.io/badge/EER%25%20on%20DeepVoice-10.24%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 30.51 on LibriSeVoc](https://img.shields.io/badge/EER%25%20on%20LibriSeVoc-30.51%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 35.07 on SONAR](https://img.shields.io/badge/EER%25%20on%20SONAR-35.07%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 13.64 on ArAD](https://img.shields.io/badge/EER%25%20on%20ArAD-13.64%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 25.44 on DECRO](https://img.shields.io/badge/EER%25%20on%20DECRO-25.44%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 5.13 on J-SPAW_LA](https://img.shields.io/badge/EER%25%20on%20J--SPAW__LA-5.13%25-yellow)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 15.31 on ODSS](https://img.shields.io/badge/EER%25%20on%20ODSS-15.31%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 21.64 on HABLA](https://img.shields.io/badge/EER%25%20on%20HABLA-21.64%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 6.07 on DFADD](https://img.shields.io/badge/EER%25%20on%20DFADD-6.07%25-yellow)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 21.75 on PyAra](https://img.shields.io/badge/EER%25%20on%20PyAra-21.75%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 8.78 on XMAD](https://img.shields.io/badge/EER%25%20on%20XMAD-8.78%25-yellow)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![EER% 25 on EmoFake_test](https://img.shields.io/badge/EER%25%20on%20EmoFake__test-25%25-lightgrey)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![SRR COMPLEMENT 12.91 on EmoSpoofTTS](https://img.shields.io/badge/SRR%20COMPLEMENT%20on%20EmoSpoofTTS-12.91%25-blue)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)
[![SRR COMPLEMENT 35.75 on LRLspoof](https://img.shields.io/badge/SRR%20COMPLEMENT%20on%20LRLspoof-35.75%25-blue)](https://huggingface.co/spaces/SpeechAntiSpoofingBenchmarks/SpeechAntiSpoofingArena?system=prosdd-asv5)

## Citation

```bibtex
@article{mahapatra2026prosdd,
  title   = {ProSDD: Learning Prosodic Representations for Speech Deepfake Detection against Expressive and Emotional Attacks},
  author  = {Mahapatra, Aurosweta and Ulgen, Ismail Rasim and Lee, Kong Aik and Andrews, Nicholas and Sisman, Berrak},
  journal = {arXiv preprint arXiv:2604.13229},
  year    = {2026}
}
```