mlsanityData to 14 Oct 2026

BEAT2AASIST model with layer fusion for ESDD 2026 Challenge

arXiv 2512.15180 ↗First posted 17 Dec 2025cs.SD cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Recent advances in audio generation have increased the risk of realistic environmental sound manipulation, motivating the ESDD 2026 Challenge as the first large-scale benchmark for Environmental Sound Deepfake Detection (ESDD). We propose BEAT2AASIST which extends BEATs-AASIST by splitting BEATs-derived representations along frequency or channel dimension and processing them with dual AASIST branches. To enrich feature representations, we incorporate top-k transformer layer fusion using concatenation, CNN-gated, and SE-gated strategies. In addition, vocoder-based data augmentation is applied to improve robustness against unseen spoofing methods. Experimental results on the official test sets demonstrate that the proposed approach achieves competitive performance across the challenge tracks.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 18 of its 21 references that are in the map. First posted on arXiv 17 Dec 2025.

Builds on

18 of 21 references.

References · 18

  1. 2020wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representationscited by 3,905
  2. 2021HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Unitscited by 2,342
  3. 2021WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processingcited by 1,657
  4. 2019SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognitioncited by 1,380
  5. 2020HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesiscited by 1,154
  6. 2022BEATs: Audio Pre-Training with Acoustic Tokenizerscited by 388
  7. 2022BigVGAN: A Universal Neural Vocoder with Large-Scale Trainingcited by 251
  8. 2021AASIST: Audio Anti-Spoofing using Integrated Spectro-Temporal Graph Attention Networkscited by 180
  9. 2021UnivNet: A Neural Vocoder with Multi-Resolution Spectrogram Discriminators for High-Fidelity Waveform Generationcited by 74
  10. 2023ADD 2023: the Second Audio Deepfake Detection Challengecited by 64
  11. 2024EAT: Self-Supervised Pre-Training with Efficient Audio Transformercited by 53
  12. 2024A Large-Scale Evaluation of Speech Foundation Modelscited by 40
6 more
  1. 2025ASVspoof 5: Design, Collection and Validation of Resources for Spoofing, Deepfake, and Adversarial Attack Detection Using Crowdsourced Speechcited by 26
  2. 2022Spoofed training data for speech spoofing countermeasure can be efficiently created using neural vocoderscited by 20
  3. 2023Can large-scale vocoded spoofed data improve speech spoofing countermeasure with a self-supervised front end?cited by 17
  4. 2025EnvSDD: Benchmarking Environmental Sound Deepfake Detectioncited by 11
  5. 2023HM-Conformer: A Conformer-based audio deepfake detection system with hierarchical pooling and multi-level classification token aggregation methodscited by 8
  6. 2025ESDD 2026: Environmental Sound Deepfake Detection Challenge Evaluation Plancited by 2

Built on by

Not cited yet.