paper-with-me

Papers

CA-MHFA: A Context-Aware Multi-Head Factorized Attentive Pooling for SSL-Based Speaker Verification

2024-09-23 · Junyi Peng, Ladislav Mošner, Lin Zhang, Oldřich Plchot, Themos Stafylakis, Lukáš Burget, Jan Černocký

Self-supervised learning (SSL) models for speaker verification (SV) have gained significant attention in recent years. However, existing SSL-based SV systems often struggle to capture local temporal dependencies and generalize across different tasks. In this paper, we propose context-aware multi-head factorized attentive pooling (CA-MHFA), a lightweight framework that incorporates contextual information from surrounding frames. CA-MHFA leverages grouped, learnable queries to effectively model contextual dependencies while maintaining efficiency by sharing keys and values across groups. Experimental results on the VoxCeleb dataset show that CA-MHFA achieves EERs of 0.42\%, 0.48\%, and 0.96\% on Vox1-O, Vox1-E, and Vox1-H, respectively, outperforming complex models like WavLM-TDNN with fewer parameters and faster convergence. Additionally, CA-MHFA demonstrates strong generalization across multiple SSL models and tasks, including emotion recognition and anti-spoofing, highlighting its robustness and versatility.

📄 PDF Abstract BibTeX arXiv:2409.15234

Code (0)

등록된 구현이 없습니다.

Tasks

Emotion RecognitionSelf-Supervised LearningSpeaker Verification

Methods 이 논문이 사용한 방법론

Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Attention 설명 없음

Similar Papers 제목 키워드 기반

Improving Out-of-Domain Audio Deepfake Detection via Layer Selection and Fusion of SSL-Based Countermeasures

2025-09-15 · Pierre Serrano, Raphaël Duroselle, Florian Angulo, Jean-François Bonastre 외 arxiv

Audio deepfake detection systems based on frozen pre-trained self-supervised learning (SSL) encoders show a high level of performance when combined with layer-weighted pooling methods, such as multi-head factorized atten…

Self-Supervised LearningAudio Deepfake Detection

Multi-Granularity Graph Pooling for Video-based Person Re-Identification

2022-09-23 · Honghu Pan, Yongyong Chen, Zhenyu He

The video-based person re-identification (ReID) aims to identify the given pedestrian video sequence across multiple non-overlapping cameras. To aggregate the temporal and spatial features of the video samples, the graph…

Node ClusteringPerson Re-IdentificationRetrievalVideo-Based Person Re-Identification+1

Factorized Hypothesis Search for Evidence-to-Taxonomy Retrieval

2026-08-06 · Linhai Ma, Ethan F. Wei, Xueqing Peng, Yan Wang 외 hf

Large-taxonomy retrieval often assumes that the input already expresses the target concept. In many settings, however, the input is indirect evidence, such as a table cell whose meaning depends on its row, column, dataty…

What Do Deepfake Speech Detectors Actually Hear?

2026-06-09 · Vojtěch Staněk, Veronika Jirmusová, Anton Firc, Kamil Malinka 외 arxiv

Deepfake speech detectors often output a single score without explaining why an audio sample is flagged, where in the signal the evidence lies, or what cues drive the decision. We propose an audio-native explainability p…

One Reflection Is Not Enough: Self-Correcting Autonomous Research via Multi-Hypothesis Failure Attribution

2026-06-30 · Jie Ma, Binfei Chu, Jie Gao, Jinlu Zhang 외 arxiv

Autonomous research agents can now draft hypotheses, write code, run experiments, and produce papers, but they remain brittle when experiments fail. Under the prevailing paradigm, failure recovery is usually delegated to…