Výsledky vyhledávání

Report

Pureformer-VC: Non-parallel One-Shot Voice Conversion with Pure Transformer Blocks and Triplet Discriminative Training

Autor: Yao, Wenhan, Xing, Zedong, Chen, Xiarun, Liu, Jia, He, Yongqiang, Wen, Weiping

One-shot voice conversion(VC) aims to change the timbre of any source speech to match that of the target speaker with only one speech sample. Existing style transfer-based VC methods relied on speech representation disentanglement and suffered from a

Externí odkaz: http://arxiv.org/abs/2409.01668

Zobrazit plný text záznamu

Report

EmoAttack: Utilizing Emotional Voice Conversion for Speech Backdoor Attacks on Deep Speech Classification Models

Autor: Yao, Wenhan, Chen, Zedong XingXiarun, Liu, Jia, He, yongqiang, Wen, Weiping

Deep speech classification tasks, mainly including keyword spotting and speaker verification, play a crucial role in speech-based human-computer interaction. Recently, the security of these technologies has been demonstrated to be vulnerable to backd

Externí odkaz: http://arxiv.org/abs/2408.15508

Zobrazit plný text záznamu

Report

Imperceptible Rhythm Backdoor Attacks: Exploring Rhythm Transformation for Embedding Undetectable Vulnerabilities on Speech Recognition

Autor: Yao, Wenhan, Yang, Jiangkun, He, Yongqiang, Liu, Jia, Wen, Weiping

Speech recognition is an essential start ring of human-computer interaction, and recently, deep learning models have achieved excellent success in this task. However, when the model training and private data provider are always separated, some securi

Externí odkaz: http://arxiv.org/abs/2406.10932

Zobrazit plný text záznamu

Report

A Point-Neighborhood Learning Framework for Nasal Endoscope Image Segmentation

Autor: Jie, Pengyu, Liu, Wanquan, Gao, Chenqiang, Wen, Yihui, He, Rui, Wen, Weiping, Li, Pengcheng, Zhang, Jintao, Meng, Deyu

Lesion segmentation on nasal endoscopic images is challenging due to its complex lesion features. Fully-supervised deep learning methods achieve promising performance with pixel-level annotations but impose a significant annotation burden on experts.

Externí odkaz: http://arxiv.org/abs/2405.20044

Zobrazit plný text záznamu