Leveraging Non-Causal Knowledge via Cross-Network Knowledge Distillation for Real-Time Speech Enhancement

Park, Hyun Joon; Shin, Wooseok; Kim, Jin Sob; Han, Sung Won

doi:10.1109/lsp.2024.3388956

Cited by 2 publications

(1 citation statement)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Direction-of-arrival (DOA) estimation plays a crucial role in various applications such as speech enhancement, speech separation, and speech recognition [1][2][3][4][5][6]. Accurate DOA estimation enables robust audio signal processing in reverberant and noise environments.…”

Section: Introductionmentioning

confidence: 99%

Robust DOA Estimation Using Multi-Scale Fusion Network with Attention Mask

Yan,

Huang

2024

Applied Sciences

View full text Add to dashboard Cite

To overcome the limitations of traditional methods in reverberant and noisy environments, a robust multi-scale fusion neural network with attention mask is designed to improve direction-of-arrival (DOA) estimation accuracy for acoustic sources. It combines the benefits of deep learning and complex-valued operations to effectively deal with the interference of reverberation and noise in speech signals. The unique properties of complex-valued signals are exploited to fully capture inherent features and rich information is preserved in the complex field. An attention mask module is designed to generate distinct masks for selectively focusing and masking based on the input. After that, the multi-scale fusion block efficiently captures multi-scale spatial features by stacking complex-valued convolutional layers with small size kernels, and reduces the module complexity through special branching operations. Experimental results demonstrate that the model achieves significant improvements over other methods for speaker localization in reverberant and noisy environments. It provides a new solution for DOA estimation for acoustic sources in different scenarios, which has significant theoretical and practical implications.

show abstract

Section: Introductionmentioning

confidence: 99%

Robust DOA Estimation Using Multi-Scale Fusion Network with Attention Mask

Yan,

Huang

2024

Applied Sciences

View full text Add to dashboard Cite

show abstract

Collaborative forecasting management model for multi‐energy microgrid considering load response characterization

Bao,

Sun,

Peng

et al. 2024

IET Renewable Power Gen

View full text Add to dashboard Cite

Multi‐energy microgrids (MEMG) have become an effective means of integrated energy management due to their unique advantages, including area independence, diverse supply, flexibility, and efficiency. However, the uncertain deviation of the renewable energy generators (REGs) output and the uncertain deviation of the multiple energy load response cumulatively lead to the deterioration of the MEMG model performance. To address these issues, this article proposes a cooperative forecasting management model for MEMG that considers multiple uncertainties and load response knowledge characterization. The model combines a multi‐energy load prediction model with a management model based on deep reinforcement learning. It proposes multiple iterations of data, fits the dynamic environment of MEMG by continuously improving the long short‐term memory (LSTM) neural network based on knowledge distillation (KD) architecture, and then optimizes the MEMG state space by considering the knowledge of load response characteristics, Furthermore, it combines multi‐agent deep deterministic policy gradient (MADDPG) with horizontal federated (hF) learning to co‐train multi‐MEMG, addressing the issues of training efficiency during co‐training. Finally, the validity of the proposed model is demonstrated by an arithmetic example.

show abstract

Leveraging Non-Causal Knowledge via Cross-Network Knowledge Distillation for Real-Time Speech Enhancement

Cited by 2 publications

References 22 publications

Robust DOA Estimation Using Multi-Scale Fusion Network with Attention Mask

Robust DOA Estimation Using Multi-Scale Fusion Network with Attention Mask

Collaborative forecasting management model for multi‐energy microgrid considering load response characterization

Contact Info

Product

Resources

About