mlsanityData to 7 Oct 2026

MFA: TDNN with Multi-scale Frequency-channel Attention for Text-independent Speaker Verification with Short Utterances

arXiv 2202.01624 ↗First posted 3 Feb 2022cs.SD cs.CL eess.AScited by 12

Not on the map. Why?

Abstract (arXiv)

The time delay neural network (TDNN) represents one of the state-of-the-art of neural solutions to text-independent speaker verification. However, they require a large number of filters to capture the speaker characteristics at any local frequency region. In addition, the performance of such systems may degrade under short utterance scenarios. To address these issues, we propose a multi-scale frequency-channel attention (MFA), where we characterize speakers at different scales through a novel dual-path design which consists of a convolutional neural network and TDNN. We evaluate the proposed MFA on the VoxCeleb database and observe that the proposed framework with MFA can achieve state-of-the-art performance while reducing parameters and computation complexity. Further, the MFA mechanism is found to be effective for speaker verification with short test utterances.

Key facts. Cited by 12 papers in the map (data through 7 Oct 2026). Builds on 17 of its 32 references that are in the map. First posted on arXiv 3 Feb 2022.

Builds on

17 of 32 references.

References · 17

  1. 2017Squeeze-and-Excitation Networkscited by 4,763
  2. 2019SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognitioncited by 1,355
  3. 2018VoxCeleb2: Deep Speaker Recognitioncited by 862
  4. 2017VoxCeleb: a large-scale speaker identification datasetcited by 714
  5. 2015Cyclical Learning Rates for Training Neural Networkscited by 627
  6. 2020ECAPA-TDNN: Emphasized Channel Attention, Propagation and Aggregation in TDNN Based Speaker Verificationcited by 606
  7. 2019ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networkscited by 572
  8. 2019Res2Net: A New Multi-scale Backbone Architecturecited by 468
  9. 2020In defence of metric learning for speaker recognitioncited by 123
  10. 2017Deep Speaker: an End-to-End Neural Speaker Embedding Systemcited by 56
  11. 2020Audio-visual Speaker Recognition with a Cross-modal Discriminative Networkcited by 24
  12. 2020The ins and outs of speaker recognition: lessons from VoxSRC 2020cited by 21
5 more
  1. 2020ResNeXt and Res2Net Structures for Speaker Verificationcited by 11
  2. 2021Integrating Frequency Translational Invariance in TDNNs and Frequency Positional Information in 2D ResNets to Enhance Speaker Verificationcited by 11
  3. 2021Xi-Vector Embedding for Speaker Recognitioncited by 8
  4. 2020Speaker-Utterance Dual Attention for Speaker and Utterance Verificationcited by 5
  5. 2021PL-EESR: Perceptual Loss Based END-TO-END Robust Speaker Representation Extractioncited by 2

Built on by

Cited by year

Citing papers per year: 2022: 1; 2023: 2; 2024: 6; 2025: 3.

12 papers.

  1. 2025Nes2Net: A Lightweight Nested Architecture for Foundation Model Driven Speech Anti-spoofingcited by 17
  2. 2022Frequency and Multi-Scale Selective Kernel Attention for Speaker Verificationcited by 10
  3. 2023Disentangling Voice and Content with Self-Supervision for Speaker Recognitioncited by 10
  4. 2024How Do Neural Spoofing Countermeasures Detect Partially Spoofed Audio?cited by 10
  5. 2024Speech Foundation Model Ensembles for the Controlled Singing Voice Deepfake Detection (CtrSVDD) Challenge 2024cited by 7
  6. 2024Towards Quantifying and Reducing Language Mismatch Effects in Cross-Lingual Speech Anti-Spoofingcited by 7
  7. 2024HiddenSpeaker: Generate Imperceptible Unlearnable Audios for Speaker Verification Systemcited by 4
  8. 2023DS-TDNN: Dual-stream Time-delay Neural Network with Global-aware Filter for Speaker Verificationcited by 1
  9. 2024MR-RawNet: Speaker verification system with multiple temporal resolutions for variable duration utterances using raw waveformscited by 1
  10. 2025Capturing More: Learning Multi-Domain Representations for Robust Online Handwriting Verificationcited by 1
  11. 2025Interpolating Speaker Identities in Embedding Space for Data Expansioncited by 1
  12. 2024Cosine Scoring with Uncertainty for Neural Speaker Embeddingcited by 0