Gradient weighting for speaker verification in extremely low Signal-to-Noise Ratio
On the map: Speech and audio language models → Speaker recognition and diarization
Abstract (arXiv)
Speaker verification is hampered by background noise, particularly at extremely low Signal-to-Noise Ratio (SNR) under 0 dB. It is difficult to suppress noise without introducing unwanted artifacts, which adversely affects speaker verification. We proposed the mechanism called Gradient Weighting (Grad-W), which dynamically identifies and reduces artifact noise during prediction. The mechanism is based on the property that the gradient indicates which parts of the input the model is paying attention to. Specifically, when the speaker network focuses on a region in the denoised utterance but not on the clean counterpart, we consider it artifact noise and assign higher weights for this region during optimization of enhancement. We validate it by training an enhancement model and testing the enhanced utterance on speaker verification. The experimental results show that our approach effectively reduces artifact noise, improving speaker verification across various SNR levels.
Key facts. No citing papers in the map yet (data through 19 Oct 2026). Placed in Speaker recognition and diarization, one of 11 directions in Speech and audio language models. Builds on 22 of its 29 references that are in the map. First posted on arXiv 5 Jan 2024.
Builds on
22 of 29 references.
Closest ancestors · 7
Same direction, newest first
- 2023Disentangling Voice and Content with Self-Supervision for Speaker Recognition2 Oct
- 2023Visualizing data augmentation in deep speaker recognition25 May
- 2022Wespeaker: A Research and Production oriented Speaker Embedding Learning Toolkit31 Oct
- 2022Extended U-Net for Speaker Verification in Noisy Environments27 Jun
- 2022MFA: TDNN with Multi-scale Frequency-channel Attention for Text-independent Speaker Verification with Short Utterances3 Feb
- 2021PL-EESR: Perceptual Loss Based END-TO-END Robust Speaker Representation Extraction3 Oct
- 2021Towards Robust Speaker Verification with Target Speaker Enhancement16 Mar
Other references · 15
- 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,762
- 2017Axiomatic Attribution for Deep Networkscited by 3,407
- 2018VoxCeleb2: Deep Speaker Recognitioncited by 1,166
- 2015MUSAN: A Music, Speech, and Noise Corpuscited by 694
- 2023LLM-Pruner: On the Structural Pruning of Large Language Modelscited by 660
- 2023DeepCache: Accelerating Diffusion Models for Freecited by 370
- 2017Real Time Image Saliency for Black Box Classifierscited by 247
- 2020The INTERSPEECH 2020 Deep Noise Suppression Challenge: Datasets, Subjective Testing Framework, and Challenge Resultscited by 230
- 2022Deep Model Reassemblycited by 79
- 2020Proper Network Interpretability Helps Adversarial Robustness in Classificationcited by 42
- 2019VoiceID Loss: Speech Enhancement for Speaker Verificationcited by 27
- 2019Feature Enhancement with Deep Feature Losses for Speaker Verificationcited by 11
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2022Adversarial Reweighting for Speaker Verification Fairness
- 2025A Framework for Robust Speaker Verification in Highly Noisy Environments Leveraging Both Noisy and Enhanced Audio
- 2025Rethinking Leveraging Pre-Trained Multi-Layer Representations for Speaker Verification
- 2025ParaNoise-SV: Integrated Approach for Noise-Robust Speaker Verification with Parallel Joint Learning of Speech Enhancement and Noise Extraction
- 2021A Speaker Verification Backend with Robust Performance across Conditions
- 2023Diff-SV: A Unified Hierarchical Framework for Noise-Robust Speaker Verification Using Score-Based Diffusion Probabilistic Models
- 2026What and When to Learn: CURriculum Ranking Loss for Large-Scale Speaker Verification
- 2025SVeritas: Benchmark for Robust Speaker Verification under Diverse Conditions