Weiqi Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Weiqi Zhang has a total of 15 co-authors. Here are the top co-authors by number of shared publications:
- Ji Xi
School of Computer Information Engineering, Changzhou Institute of Technology, Changzhou 213022, China
6 shared publications - Li Zhao
School of Information Science and Engineering, Southeast University, Nanjing 210096, China
4 shared publications - Zhe Xu
School of Computer Information and Engineering, Changzhou Institute of Technology, Changzhou 213032, China
3 shared publications - Zhengwang Xia
School of Computer Information Engineering, Changzhou Institute of Technology, Changzhou 213022, China
3 shared publications - Yue Xie
School of Communication Engineering, Nanjing Institute of Technology, Nanjing 211167, China
2 shared publications - Fukun Chen
School of Computer and Artificial Intelligence, Liaoning Normal University, Dalian 116029, China
1 shared publication
Latest Publications
- Open Access
- 16 November 2026
- Published by MDPI AG in Electronics
- Vol. 15 (22) , 5334
Diagnosing Modality Reliability in Sparse E-Commerce Recommendation: A White-Box Routing Study
Multimodal recommenders usually assume that fusing collaborative, visual, and textual signals improves accuracy. We treat this as a diagnostic question rather than an accuracy contest. Under interaction-level leave-one-out evaluation on three sparse Amazon categories, three findings emerge. First, text is the only modality that transfers; removing it lowers AUC by –, while removing the visual stream changes AUC by at most and removing collaborative ID features leaves it unchanged. Second, the apparent value of collaborative features is protocol-dependent. Full models reach AUC – under random splitting but only – under leave-one-out. The gap vanishes under a user-disjoint split, which is consistent with identity-side shortcuts rather than transferable signal. Third, white-box routing agrees with modality ablations at the dataset level, but is not seed-stable enough to serve as a per-instance explanation. We introduce counterfactual modality masking, which yields signed per-item modality attributions and shows that a modality can help on average while acting as noise for a specific item; gradient-based attribution, which aggregates absolute magnitudes, cannot express this sign. Against recent 2025–2026 multimodal models re-implemented under the identical protocol, LCFC is significantly more accurate than OrthoRec and statistically indistinguishable from I3MRec. Its contribution is not a leaderboard claim but a leakage-aware evaluation protocol together with an auditable diagnostic instrument. All deltas carry five-seed dispersion and paired significance tests.- 16 November 2026
- Published by MDPI AG in Electronics
- Vol. 15 (22) , 5334
Open AccessDiagnosing Modality Reliability in Sparse E-Commerce Recommendation: A White-Box Routing Study
Multimodal recommenders usually assume that fusing collaborative, visual, and textual signals improves accuracy. We treat this as a diagnostic question rather than an accuracy contest. Under interaction-level leave-one-out evaluation on three sparse Amazon categories, three findings emerge. First, text is the only modality that transfers; removing it lowers AUC by –, while removing the visual stream changes AUC by at most and removing collaborative ID features leaves it unchanged. Second, the apparent value of collaborative features is protocol-dependent. Full models reach AUC – under random splitting but only – under leave-one-out. The gap vanishes under a user-disjoint split, which is consistent with identity-side shortcuts rather than transferable signal. Third, white-box routing agrees with modality ablations at the dataset level, but is not seed-stable enough to serve as a per-instance explanation. We introduce counterfactual modality masking, which yields signed per-item modality attributions and shows that a modality can help on average while acting as noise for a specific item; gradient-based attribution, which aggregates absolute magnitudes, cannot express this sign. Against recent 2025–2026 multimodal models re-implemented under the identical protocol, LCFC is significantly more accurate than OrthoRec and statistically indistinguishable from I3MRec. Its contribution is not a leaderboard claim but a leakage-aware evaluation protocol together with an auditable diagnostic instrument. All deltas carry five-seed dispersion and paired significance tests. - Open Access
- 14 October 2025
- Published by MDPI AG in Electronics
- Vol. 14 (20) , 4019
Diffusion-Based Model for Audio Steganography
Audio steganography exploits redundancies in the human auditory system to conceal secret information within cover audio, ensuring that the hidden data remains undetectable during normal listening. However, recent research shows that current audio steganography techniques are vulnerable to detection by deep learning-based steganalyzers, which analyze the high-dimensional features of stego audio for classification. While deep learning-based steganography has been extensively studied for image covers, its application to audio remains underexplored, particularly in achieving robust embedding and extraction with minimal perceptual distortion. We propose a diffusion-based audio steganography model comprising two primary modules: (i) a diffusion-based embedding module that autonomously integrates secret messages into cover audio while preserving high perceptual quality and (ii) a corresponding diffusion-based extraction module that accurately recovers the embedded data. The framework supports both pre-existing cover audio and the generation of high-quality steganographic cover audio with superior perceptual quality for message embedding. After training, the model achieves state-of-the-art performance in terms of embedding capacity and resistance to detection by deep learning steganalyzers. The experimental results demonstrate that our diffusion-based approach significantly outperforms existing methods across varying embedding rates, yielding stego audio with superior auditory quality and lower detectability.
- 14 October 2025
- Published by MDPI AG in Electronics
- Vol. 14 (20) , 4019
Open AccessDiffusion-Based Model for Audio Steganography
Audio steganography exploits redundancies in the human auditory system to conceal secret information within cover audio, ensuring that the hidden data remains undetectable during normal listening. However, recent research shows that current audio steganography techniques are vulnerable to detection by deep learning-based steganalyzers, which analyze the high-dimensional features of stego audio for classification. While deep learning-based steganography has been extensively studied for image covers, its application to audio remains underexplored, particularly in achieving robust embedding and extraction with minimal perceptual distortion. We propose a diffusion-based audio steganography model comprising two primary modules: (i) a diffusion-based embedding module that autonomously integrates secret messages into cover audio while preserving high perceptual quality and (ii) a corresponding diffusion-based extraction module that accurately recovers the embedded data. The framework supports both pre-existing cover audio and the generation of high-quality steganographic cover audio with superior perceptual quality for message embedding. After training, the model achieves state-of-the-art performance in terms of embedding capacity and resistance to detection by deep learning steganalyzers. The experimental results demonstrate that our diffusion-based approach significantly outperforms existing methods across varying embedding rates, yielding stego audio with superior auditory quality and lower detectability.
- Open Access
- 10 September 2025
- Published by MDPI AG in Brain Sciences
- Vol. 15 (9) , 975
Brain Connectivity Estimation Network for the Identification of Dementia
Objectives: The brain network serves as a reliable tool for diagnosing neurological disorders. However, the current modeling algorithms for brain networks often rely on several assumptions regarding the interactions between brain regions, which can be inaccurate. For instance, some studies assume linear relationships among brain regions. Additionally, some research suggests that certain brain regions do not significantly influence outcomes when assessing directional influence between paired regions. Methods: To address this issue, we introduced a novel method for modeling brain connectivity structures that estimates interactions among regions from a different perspective. More importantly, this method considers all the relevant brain regions during evaluation rather than isolating individual relationships. Results: To validate its effectiveness, we conducted extensive experiments using publicly available datasets. The proposed method achieved superior performance across all tasks. Conclusions: The results demonstrate that our method not only excels in identifying various brain disorders but also uncovers new biomarkers, providing fresh insights into neurological disorder research.
- 10 September 2025
- Published by MDPI AG in Brain Sciences
- Vol. 15 (9) , 975
Open AccessBrain Connectivity Estimation Network for the Identification of Dementia
Objectives: The brain network serves as a reliable tool for diagnosing neurological disorders. However, the current modeling algorithms for brain networks often rely on several assumptions regarding the interactions between brain regions, which can be inaccurate. For instance, some studies assume linear relationships among brain regions. Additionally, some research suggests that certain brain regions do not significantly influence outcomes when assessing directional influence between paired regions. Methods: To address this issue, we introduced a novel method for modeling brain connectivity structures that estimates interactions among regions from a different perspective. More importantly, this method considers all the relevant brain regions during evaluation rather than isolating individual relationships. Results: To validate its effectiveness, we conducted extensive experiments using publicly available datasets. The proposed method achieved superior performance across all tasks. Conclusions: The results demonstrate that our method not only excels in identifying various brain disorders but also uncovers new biomarkers, providing fresh insights into neurological disorder research.
- Open Access
A Semantic Behavioral Sequence-Based Approach to Trajectory Privacy Protection
Trajectory data contain numerous sensitive attributes. Unauthorized disclosure of precise user trajectory information generates persistent privacy and security concerns that significantly impact daily life. Most existing trajectory privacy protection schemes focus on geographic trajectories while neglecting the critical importance of semantic trajectories, resulting in ongoing privacy vulnerabilities. To address this limitation, we propose the Semantic Behavior Sequence-based Trajectory Privacy Protection method (SBS-TPP). Our approach integrates short-term and long-term behavioral patterns within a user behavior modeling layer to identify user preferences. A dual-model framework (geographic and semantic) generates noise-injected trajectories with maximized noise potential. This methodology applies symmetric noise addition to both geographic trajectory fragments and semantic trajectory segments, optimizing trajectory data utility while ensuring robust protection of sensitive information. The SBS-TPP framework operates in the following two phases: firstly, behavior modeling, which comprises interest extraction from behavioral trajectory sequences, and secondly, noise generation, which creates synthetic noise locations with maximal semantic expectation from original locations, yielding privacy-enhanced trajectories for publication. Experimental results demonstrate that our interest extraction model achieves 93.7% accuracy while maintaining 81.6% data utility under strict privacy guarantees. The proposed method significantly enhances data usability and enables effective recommendation services without compromising user privacy or security.
Open AccessA Semantic Behavioral Sequence-Based Approach to Trajectory Privacy Protection
Trajectory data contain numerous sensitive attributes. Unauthorized disclosure of precise user trajectory information generates persistent privacy and security concerns that significantly impact daily life. Most existing trajectory privacy protection schemes focus on geographic trajectories while neglecting the critical importance of semantic trajectories, resulting in ongoing privacy vulnerabilities. To address this limitation, we propose the Semantic Behavior Sequence-based Trajectory Privacy Protection method (SBS-TPP). Our approach integrates short-term and long-term behavioral patterns within a user behavior modeling layer to identify user preferences. A dual-model framework (geographic and semantic) generates noise-injected trajectories with maximized noise potential. This methodology applies symmetric noise addition to both geographic trajectory fragments and semantic trajectory segments, optimizing trajectory data utility while ensuring robust protection of sensitive information. The SBS-TPP framework operates in the following two phases: firstly, behavior modeling, which comprises interest extraction from behavioral trajectory sequences, and secondly, noise generation, which creates synthetic noise locations with maximal semantic expectation from original locations, yielding privacy-enhanced trajectories for publication. Experimental results demonstrate that our interest extraction model achieves 93.7% accuracy while maintaining 81.6% data utility under strict privacy guarantees. The proposed method significantly enhances data usability and enables effective recommendation services without compromising user privacy or security.
- Open Access
- 10 February 2025
- Published by MDPI AG in Electronics
- Vol. 14 (4) , 681
Speech Enhancement Algorithm Based on Microphone Array and Multi-Channel Parallel GRU-CNN Network
This paper presents an improved speech enhancement algorithm based on microphone arrays to improve speech enhancement performance in complex settings. The algorithm’s model consists of two key components: the feature extraction module and the speech enhancement module. The feature extraction module processes the speech amplitude spectral features derived from STFT (short-time Fourier transform). It employs parallel GRU-CNN (Gated Recurrent Units and CNN Convolutional Neural Network) structures to capture unique channel information, and skip connections are utilized to enhance the model’s convergence speed. The speech enhancement module focuses on obtaining cross-channel spatial information. By introducing an attention mechanism and applying a global hybrid pooling strategy, it reduces feature loss. This strategy dynamically assigns weights to each channel, emphasizing features that are most beneficial for speech signal restoration. Experimental results on the CHIME3 dataset show that the proposed model effectively suppresses diverse types of noise and outperforms other algorithms in improving speech quality and comprehension.
- 10 February 2025
- Published by MDPI AG in Electronics
- Vol. 14 (4) , 681
Open AccessSpeech Enhancement Algorithm Based on Microphone Array and Multi-Channel Parallel GRU-CNN Network
This paper presents an improved speech enhancement algorithm based on microphone arrays to improve speech enhancement performance in complex settings. The algorithm’s model consists of two key components: the feature extraction module and the speech enhancement module. The feature extraction module processes the speech amplitude spectral features derived from STFT (short-time Fourier transform). It employs parallel GRU-CNN (Gated Recurrent Units and CNN Convolutional Neural Network) structures to capture unique channel information, and skip connections are utilized to enhance the model’s convergence speed. The speech enhancement module focuses on obtaining cross-channel spatial information. By introducing an attention mechanism and applying a global hybrid pooling strategy, it reduces feature loss. This strategy dynamically assigns weights to each channel, emphasizing features that are most beneficial for speech signal restoration. Experimental results on the CHIME3 dataset show that the proposed model effectively suppresses diverse types of noise and outperforms other algorithms in improving speech quality and comprehension.
- Open Access
- 9 November 2024
- Published by MDPI AG in Electronics
- Vol. 13 (22) , 4394
Speech Enhancement Algorithm Based on Microphone Array and Lightweight CRN for Hearing Aid
To address the performance and computational complexity issues in speech enhancement for hearing aids, a speech enhancement algorithm based on a microphone array and a lightweight two-stage convolutional recurrent network (CRN) is proposed. The algorithm consists of two main modules: a beamforming module and a post-filtering module. The beamforming module utilizes directional features and a complex time-frequency long short-term memory (CFT-LSTM) network to extract local representations and perform spatial filtering. The post-filtering module uses analogous encoding and two symmetric decoding structures, with stacked CFT-LSTM blocks in between. It further reduces residual noise and improves filtering performance by passing spatial information through an inter-channel masking module. Experimental results show that this algorithm outperforms existing methods on the generated hearing aid dataset and the CHIME-3 dataset, with fewer parameters and lower model complexity, making it suitable for hearing aid scenarios with limited computational resources.
- 9 November 2024
- Published by MDPI AG in Electronics
- Vol. 13 (22) , 4394
Open AccessSpeech Enhancement Algorithm Based on Microphone Array and Lightweight CRN for Hearing Aid
To address the performance and computational complexity issues in speech enhancement for hearing aids, a speech enhancement algorithm based on a microphone array and a lightweight two-stage convolutional recurrent network (CRN) is proposed. The algorithm consists of two main modules: a beamforming module and a post-filtering module. The beamforming module utilizes directional features and a complex time-frequency long short-term memory (CFT-LSTM) network to extract local representations and perform spatial filtering. The post-filtering module uses analogous encoding and two symmetric decoding structures, with stacked CFT-LSTM blocks in between. It further reduces residual noise and improves filtering performance by passing spatial information through an inter-channel masking module. Experimental results show that this algorithm outperforms existing methods on the generated hearing aid dataset and the CHIME-3 dataset, with fewer parameters and lower model complexity, making it suitable for hearing aid scenarios with limited computational resources.
- Open Access
Trajectory Privacy-Protection Mechanism Based on Multidimensional Spatial–Temporal Prediction
The popularity of global GPS location services and location-enabled personal terminal applications has contributed to the rapid growth of location-based social networks. Users can access social networks at anytime and anywhere to obtain services in the relevant location. While accessing services is convenient, there is a potential risk of leaking users’ private information. In data processing, the discovery of issues and the generation of optimal solutions constitute a symmetrical process. Therefore, this paper proposes a symmetry–trajectory differential privacy-protection mechanism based on multi-dimensional prediction (TPPM-MP). Firstly, the temporal attention mechanism is designed to extract spatiotemporal features of trajectories from different spatiotemporal dimensions and perform trajectory-sensitive prediction. Secondly, class-prevalence-based weights are assigned to sensitive regions. Finally, the privacy budget is assigned based on the sensitive weights, and noise conforming to localized differential privacy is added. Validated on real datasets, the proposed method in this paper enhanced usability by 22% and 37% on the same dataset compared with other methods mentioned, while providing equivalent privacy protection.
Open AccessTrajectory Privacy-Protection Mechanism Based on Multidimensional Spatial–Temporal Prediction
The popularity of global GPS location services and location-enabled personal terminal applications has contributed to the rapid growth of location-based social networks. Users can access social networks at anytime and anywhere to obtain services in the relevant location. While accessing services is convenient, there is a potential risk of leaking users’ private information. In data processing, the discovery of issues and the generation of optimal solutions constitute a symmetrical process. Therefore, this paper proposes a symmetry–trajectory differential privacy-protection mechanism based on multi-dimensional prediction (TPPM-MP). Firstly, the temporal attention mechanism is designed to extract spatiotemporal features of trajectories from different spatiotemporal dimensions and perform trajectory-sensitive prediction. Secondly, class-prevalence-based weights are assigned to sensitive regions. Finally, the privacy budget is assigned based on the sensitive weights, and noise conforming to localized differential privacy is added. Validated on real datasets, the proposed method in this paper enhanced usability by 22% and 37% on the same dataset compared with other methods mentioned, while providing equivalent privacy protection.