Liang Tao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Liang Tao has a total of 86 co-authors. Here are the top co-authors by number of shared publications:
- Huabin Wang
Anhui Provincial Key Laboratory of Multimodal Cognitive Computation, School of Computer Science and Technology, Anhui University, Hefei, China
19 shared publications - Jian Zhou
Anhui University
14 shared publications - Hon Keung Kwan
Department of Electrical and Computer Engineering, University of Windsor, Windsor, ON, Canada
8 shared publications - Fei Liu
School of Engineering, Monash University Malaysia, Kuala Lumpur, Malaysia
3 shared publications - Yonglin Chen
Anhui University
3 shared publications - Bowen Xu
State Key Laboratory of Industrial Control Technology and the College of Control Science and Engineering, Zhejiang University, Hangzhou, China
2 shared publications
Latest Publications
- 14 February 2026
- Published by Springer Nature in Applied Intelligence
- Vol. 56 (3) , 88
Multimodal hand feature recognition base on modality-related feature interaction
- 14 February 2026
- Published by Springer Nature in Applied Intelligence
- Vol. 56 (3) , 88
Multimodal hand feature recognition base on modality-related feature interaction
- 9 September 2025
- Published by Springer Nature in Circuits, Systems, and Signal Processing
- Vol. 45 (3) , 1979-1999
Multi-path Processing Structure for Multi-scale Feature Fusion in Speech Separation Transformer
- 9 September 2025
- Published by Springer Nature in Circuits, Systems, and Signal Processing
- Vol. 45 (3) , 1979-1999
Multi-path Processing Structure for Multi-scale Feature Fusion in Speech Separation Transformer
- 4 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Emerging Topics in Computing
- Vol. 13 (3) , 1200-1214
The Cancelable Multimodal Template Protection Algorithm Based on Random Index
Current multimodal template protection methods typically require encryption or transformation of the original biometric features. However, these operations carry certain risks, as attackers may reverse-engineer or decrypt the protected multimodal templates to retrieve partial or complete information about the original templates, leading to the leakage of the original biometric features. To address this issue, we propose a cancelable multimodal template protection method based on random indexing. First, hash functions are used to generate integer sequences as index values, which are then employed to create single-modal cancelable templates using random binary vectors. Second, the single-modal cancelable templates are used as indices for random binary sequences, which locate the corresponding template information and are filled into the fusion cancelable template at the respective positions, achieving template fusion. The resulting template is unrelated to the original biometric features. Finally, without directly storing the binary factor sequences, an XOR operation is performed on the extended biometric feature vectors and random binary sequences to generate the encoded key. Experimental results demonstrate that the proposed method significantly enhances performance on the FVC2002DB1 fingerprint, MMCBNU_6000 finger-vein, and NUPT_FPV databases, while also satisfying the standards for cancelable biometric feature design. We also analyze four privacy and security attacks against this scheme.
- 4 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Emerging Topics in Computing
- Vol. 13 (3) , 1200-1214
The Cancelable Multimodal Template Protection Algorithm Based on Random Index
Current multimodal template protection methods typically require encryption or transformation of the original biometric features. However, these operations carry certain risks, as attackers may reverse-engineer or decrypt the protected multimodal templates to retrieve partial or complete information about the original templates, leading to the leakage of the original biometric features. To address this issue, we propose a cancelable multimodal template protection method based on random indexing. First, hash functions are used to generate integer sequences as index values, which are then employed to create single-modal cancelable templates using random binary vectors. Second, the single-modal cancelable templates are used as indices for random binary sequences, which locate the corresponding template information and are filled into the fusion cancelable template at the respective positions, achieving template fusion. The resulting template is unrelated to the original biometric features. Finally, without directly storing the binary factor sequences, an XOR operation is performed on the extended biometric feature vectors and random binary sequences to generate the encoded key. Experimental results demonstrate that the proposed method significantly enhances performance on the FVC2002DB1 fingerprint, MMCBNU_6000 finger-vein, and NUPT_FPV databases, while also satisfying the standards for cancelable biometric feature design. We also analyze four privacy and security attacks against this scheme.
- 9 March 2025
- Published by Springer Nature
- p. 72-81
Adversarial Attacks on Facial Images Based on Attribute-Conditioned High-Camouflage Editing
- 9 March 2025
- Published by Springer Nature
- p. 72-81
Adversarial Attacks on Facial Images Based on Attribute-Conditioned High-Camouflage Editing
- 17 February 2025
- Published by Springer Nature in Applied Intelligence
- Vol. 55 (7) , 1-23
Learning sparse filters-based convolutional networks without offline training for robust visual tracking
- 17 February 2025
- Published by Springer Nature in Applied Intelligence
- Vol. 55 (7) , 1-23
Learning sparse filters-based convolutional networks without offline training for robust visual tracking
- 11 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Affective Computing
- Vol. 16 (1) , 68-82
Controllable Multi-Speaker Emotional Speech Synthesis With an Emotion Representation of High Generalization Capability
The aim of multi-speaker emotional speech synthesis is to generate speech for a designated speaker in a desired emotional state. The task is challenging due to the presence of speech variations, such as noise, content, and timbre, which can obstruct emotion extraction and transfer. This paper proposes a new approach to performing multi-speaker emotional speech synthesis. The proposed method, which is based on a seq2seq synthesizer, integrates emotion embedding as a conditioned variable to convey exact emotional information from reference audio to the synthesized speech. To dig emotion representation capability, we utilize a three-dimensional acoustic feature as input. And an emotion generalization module with adaptive instance normalization (AdaIN) is proposed to obtain emotion embedding with high generalization ability, which also results in improved controllability. The derived emotion embedding from the generalization module can be readily conditioned by affine parameters, allowing for control both the emotion category and the emotion intensity of synthesized speech. Various emotional speech synthesis experimental results of the propposed method demonstrate its state-of-the-art performance in multi-speaker emotional speech synthesis, coupled with its advantage of high emotion controllability.
- 11 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Affective Computing
- Vol. 16 (1) , 68-82
Controllable Multi-Speaker Emotional Speech Synthesis With an Emotion Representation of High Generalization Capability
The aim of multi-speaker emotional speech synthesis is to generate speech for a designated speaker in a desired emotional state. The task is challenging due to the presence of speech variations, such as noise, content, and timbre, which can obstruct emotion extraction and transfer. This paper proposes a new approach to performing multi-speaker emotional speech synthesis. The proposed method, which is based on a seq2seq synthesizer, integrates emotion embedding as a conditioned variable to convey exact emotional information from reference audio to the synthesized speech. To dig emotion representation capability, we utilize a three-dimensional acoustic feature as input. And an emotion generalization module with adaptive instance normalization (AdaIN) is proposed to obtain emotion embedding with high generalization ability, which also results in improved controllability. The derived emotion embedding from the generalization module can be readily conditioned by affine parameters, allowing for control both the emotion category and the emotion intensity of synthesized speech. Various emotional speech synthesis experimental results of the propposed method demonstrate its state-of-the-art performance in multi-speaker emotional speech synthesis, coupled with its advantage of high emotion controllability.
- 3 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 31 (10709908) , 1580-1584
Multi-Level Information Aggregation Based Graph Attention Networks Towards Fake Speech Detection
It is widely acknowledged that distinguishing genuine speech from spoofed speech encompasses various subbands and temporal segments within speech signals. However, prevailing spoofing detection methods tend to oversimplify the relationships between these cues by employing linear models. In this paper, we introduce a multi-level information aggregation Graph Attention Networks (MiaGATs) to generate highly discriminative features for fake speech detection (FSD). In MiaGATs, each subband and temporal segment of a speech signal is represented as distinct nodes. MiaGATs incorporates channel information aggregation within each node to effectively harness the unique spectral and temporal characteristics during the feature encoding stage. In particular, MiaGATs address the interactions between nodes through indirect node aggregation and integrates both indirect and direct node aggregation by max-pooling operation. Experimental results on ASVspoof2019 and ASVspoof2021 LA databases show significant relative improvement compared to the current state-of-the-art. In comparison to the leading integrated spectro-temporal graph attention networks, MiaGATs gains an impressive performance improvement in various conditions, underscoring MiaGATs's position as a new benchmark in spoofing detection performance.
- 3 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 31 (10709908) , 1580-1584
Multi-Level Information Aggregation Based Graph Attention Networks Towards Fake Speech Detection
It is widely acknowledged that distinguishing genuine speech from spoofed speech encompasses various subbands and temporal segments within speech signals. However, prevailing spoofing detection methods tend to oversimplify the relationships between these cues by employing linear models. In this paper, we introduce a multi-level information aggregation Graph Attention Networks (MiaGATs) to generate highly discriminative features for fake speech detection (FSD). In MiaGATs, each subband and temporal segment of a speech signal is represented as distinct nodes. MiaGATs incorporates channel information aggregation within each node to effectively harness the unique spectral and temporal characteristics during the feature encoding stage. In particular, MiaGATs address the interactions between nodes through indirect node aggregation and integrates both indirect and direct node aggregation by max-pooling operation. Experimental results on ASVspoof2019 and ASVspoof2021 LA databases show significant relative improvement compared to the current state-of-the-art. In comparison to the leading integrated spectro-temporal graph attention networks, MiaGATs gains an impressive performance improvement in various conditions, underscoring MiaGATs's position as a new benchmark in spoofing detection performance.
- 7 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Artificial Intelligence
- Vol. 5 (9) , 4753-4764
Variable Curvature Gabor Convolution and Multibranch Structures for Finger Vein Recognition
Gabor filters are able to extract texture features from finger vein images from different directions and scales. However, manually crafted Gabor filters have problems such as relatively single direction and scale, and difficulties in parameter adjustment to adapt to specific datasets. To solve these problems, this paper proposes a neural network with a learnable variable curvature Gabor (VC-Gabor) convolutional layer. Firstly, the Gabor filter is improved by adding variable curvature to extract information about different curvature degrees in the vein curves. Secondly, the VC-Gabor filter is designed as a learnable convolutional filter, with parameters updated using neural network back-propagation. This facilitates the enrichment of learned VC-Gabor filter directions, scales, and curvatures, eliminating the need for intricate manual parameter tuning. Finally, we propose adaptive multi-branch structures for feature extraction, which are used to enhance the feature extraction capability of the model. Experimental results on publicly available datasets FV-USM and SDUMLA demonstrate that the proposed algorithm improves recognition accuracy and reduces the Equal Error Rate (EER), thereby substantiating the effectiveness of the approach.
- 7 May 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Artificial Intelligence
- Vol. 5 (9) , 4753-4764
Variable Curvature Gabor Convolution and Multibranch Structures for Finger Vein Recognition
Gabor filters are able to extract texture features from finger vein images from different directions and scales. However, manually crafted Gabor filters have problems such as relatively single direction and scale, and difficulties in parameter adjustment to adapt to specific datasets. To solve these problems, this paper proposes a neural network with a learnable variable curvature Gabor (VC-Gabor) convolutional layer. Firstly, the Gabor filter is improved by adding variable curvature to extract information about different curvature degrees in the vein curves. Secondly, the VC-Gabor filter is designed as a learnable convolutional filter, with parameters updated using neural network back-propagation. This facilitates the enrichment of learned VC-Gabor filter directions, scales, and curvatures, eliminating the need for intricate manual parameter tuning. Finally, we propose adaptive multi-branch structures for feature extraction, which are used to enhance the feature extraction capability of the model. Experimental results on publicly available datasets FV-USM and SDUMLA demonstrate that the proposed algorithm improves recognition accuracy and reduces the Equal Error Rate (EER), thereby substantiating the effectiveness of the approach.
- 23 October 2023
- Published by IOP Publishing in Physics in Medicine & Biology
- Vol. 68 (21) , 215005
EAMNet: an Alzheimer’s disease prediction model based on representation learning
Objective . Brain 18 F-FDG PET images indicate brain lesions’ metabolic status and offer the predictive potential for Alzheimer’s disease (AD). However, the complexity of extracting relevant lesion features and dealing with extraneous information in PET images poses challenges for accurate prediction. Approach . To address these issues, we propose an innovative solution called the efficient adaptive multiscale network (EAMNet) for predicting potential patient populations using positron emission tomography (PET) image slices, enabling effective intervention and treatment. Firstly, we introduce an efficient convolutional strategy to enhance the receptive field of PET images during the feature learning process, avoiding excessive extraction of fine tissue features by deep-level networks while reducing the model’s computational complexity. Secondly, we construct a channel attention module that enables the prediction model to adaptively allocate weights between different channels, compensating for the spatial noise in PET images’ impact on classification. Finally, we use skip connections to merge features from different-scale lesion information. Through visual analysis, the network constructed in this article aligns with the regions of interest of clinical doctors. Main results . Through visualization analysis, our network aligns with regions of interest identified by clinical doctors. Experimental evaluations conducted on the ADNI (Alzheimer’s Disease Neuroimaging Initiative) dataset demonstrate the outstanding classification performance of our proposed method. The accuracy rates for AD versus NC (Normal Controls), AD versus MCI (Mild Cognitive Impairment), MCI versus NC, and AD versus MCI versus NC classifications achieve 97.66%, 96.32%, 95.23%, and 95.68%, respectively. Significance . The proposed method surpasses advanced algorithms in the field, providing a hopeful advancement in accurately predicting and classifying Alzheimer’s Disease using 18 F-FDG PET images. The source code has been uploaded to https://github.com/Haoliang-D-AHU/EAMNet/tree/master .
- 23 October 2023
- Published by IOP Publishing in Physics in Medicine & Biology
- Vol. 68 (21) , 215005
EAMNet: an Alzheimer’s disease prediction model based on representation learning
Objective . Brain 18 F-FDG PET images indicate brain lesions’ metabolic status and offer the predictive potential for Alzheimer’s disease (AD). However, the complexity of extracting relevant lesion features and dealing with extraneous information in PET images poses challenges for accurate prediction. Approach . To address these issues, we propose an innovative solution called the efficient adaptive multiscale network (EAMNet) for predicting potential patient populations using positron emission tomography (PET) image slices, enabling effective intervention and treatment. Firstly, we introduce an efficient convolutional strategy to enhance the receptive field of PET images during the feature learning process, avoiding excessive extraction of fine tissue features by deep-level networks while reducing the model’s computational complexity. Secondly, we construct a channel attention module that enables the prediction model to adaptively allocate weights between different channels, compensating for the spatial noise in PET images’ impact on classification. Finally, we use skip connections to merge features from different-scale lesion information. Through visual analysis, the network constructed in this article aligns with the regions of interest of clinical doctors. Main results . Through visualization analysis, our network aligns with regions of interest identified by clinical doctors. Experimental evaluations conducted on the ADNI (Alzheimer’s Disease Neuroimaging Initiative) dataset demonstrate the outstanding classification performance of our proposed method. The accuracy rates for AD versus NC (Normal Controls), AD versus MCI (Mild Cognitive Impairment), MCI versus NC, and AD versus MCI versus NC classifications achieve 97.66%, 96.32%, 95.23%, and 95.68%, respectively. Significance . The proposed method surpasses advanced algorithms in the field, providing a hopeful advancement in accurately predicting and classifying Alzheimer’s Disease using 18 F-FDG PET images. The source code has been uploaded to https://github.com/Haoliang-D-AHU/EAMNet/tree/master .
- Open Access
- 9 March 2023
- Published by MDPI AG in Algorithms
- Vol. 16 (3) , 153
Speaker-Independent Spectral Enhancement for Bone-Conducted Speech
Because of the acoustic characteristics of bone-conducted (BC) speech, BC speech can be enhanced to better communicate in a complex environment with high noise. Existing BC speech enhancement models have weak spectral recovery capability for the high-frequency part of BC speech and have poor enhancement and robustness for the speaker-independent BC speech datasets. To improve the enhancement effect of BC speech for speaker-independent speech enhancement, we use a GANs method to establish the feature mapping between BC and air-conducted (AC) speech to recover the missing components of BC speech. In addition, the method adds the training of the spectral distance constraint model and, finally, uses the enhanced model completed by the training to reconstruct the BC speech. The experimental results show that this method is superior to the comparison methods such as CycleGAN, BLSTM, GMM, and StarGAN in terms of speaker-independent BC speech enhancement and can obtain higher subjective and objective evaluation results of enhanced BC speech.
- 9 March 2023
- Published by MDPI AG in Algorithms
- Vol. 16 (3) , 153
Open AccessSpeaker-Independent Spectral Enhancement for Bone-Conducted Speech
Because of the acoustic characteristics of bone-conducted (BC) speech, BC speech can be enhanced to better communicate in a complex environment with high noise. Existing BC speech enhancement models have weak spectral recovery capability for the high-frequency part of BC speech and have poor enhancement and robustness for the speaker-independent BC speech datasets. To improve the enhancement effect of BC speech for speaker-independent speech enhancement, we use a GANs method to establish the feature mapping between BC and air-conducted (AC) speech to recover the missing components of BC speech. In addition, the method adds the training of the spectral distance constraint model and, finally, uses the enhanced model completed by the training to reconstruct the BC speech. The experimental results show that this method is superior to the comparison methods such as CycleGAN, BLSTM, GMM, and StarGAN in terms of speaker-independent BC speech enhancement and can obtain higher subjective and objective evaluation results of enhanced BC speech.