arXiv subjects

Jianjun Hao

Publications and source records attributed to Jianjun Hao.

5 recordsLinked to original sources

arXiv2024

DMCE: Diffusion Model Channel Enhancer for Multi-User Semantic Communication Systems

To achieve continuous massive data transmission with significantly reduced data payload, the users can adopt semantic communication techniques to compress the redundant information by transmitting semantic features instead. However, current works on semantic communication mainly focus on high compression ratio, neglecting the wireless channel effects including dynamic distortion and multi-user interference, which significantly limit the fidelity of semantic communication. To address this, this paper proposes a diffusion model (DM)-based channel enhancer (DMCE) for improving the performance of multi-user semantic communication, with the DM learning the particular data distribution of channel effects on the transmitted semantic features. In the considered system model, multiple users (such as road cameras) transmit semantic features of multi-source data to a receiver by applying the joint source-channel coding (JSCC) techniques, and the receiver fuses the semantic features from multiple users to complete specific tasks. Then, we propose DMCE to enhance the channel state information (CSI) estimation for improving the restoration of the received semantic features. Finally, the fusion results at the receiver are significantly enhanced, demonstrating a robust performance even under low signal-to-noise ratio (SNR) regimes, enabling the generation of effective object segmentation images. Extensive simulation results with a traffic scenario dataset show that the proposed scheme can improve the mean Intersection over Union (mIoU) by more than 25\% at low SNR regimes, compared with the benchmark schemes.

eess.SP↗

arXiv2022

U-Former: Improving Monaural Speech Enhancement with Multi-head Self and Cross Attention

For supervised speech enhancement, contextual information is important for accurate spectral mapping. However, commonly used deep neural networks (DNNs) are limited in capturing temporal contexts. To leverage long-term contexts for tracking a target speaker, this paper treats the speech enhancement as sequence-to-sequence mapping, and propose a novel monaural speech enhancement U-net structure based on Transformer, dubbed U-Former. The key idea is to model long-term correlations and dependencies, which are crucial for accurate noisy speech modeling, through the multi-head attention mechanisms. For this purpose, U-Former incorporates multi-head attention mechanisms at two levels: 1) a multi-head self-attention module which calculate the attention map along both time- and frequency-axis to generate time and frequency sub-attention maps for leveraging global interactions between encoder features, while 2) multi-head cross-attention module which are inserted in the skip connections allows a fine recovery in the decoder by filtering out uncorrelated features. Experimental results illustrate that the U-Former obtains consistently better performance than recent models of PESQ, STOI, and SSNR scores.

eess.AS↗

arXiv2022

AMFFCN: Attentional Multi-layer Feature Fusion Convolution Network for Audio-visual Speech Enhancement

Audio-visual speech enhancement system is regarded to be one of promising solutions for isolating and enhancing speech of desired speaker. Conventional methods focus on predicting clean speech spectrum via a naive convolution neural network based encoder-decoder architecture, and these methods a) not adequate to use data fully and effectively, b) cannot process features selectively. The proposed model addresses these drawbacks, by a) applying a model that fuses audio and visual features layer by layer in encoding phase, and that feeds fused audio-visual features to each corresponding decoder layer, and more importantly, b) introducing soft threshold attention into the model to select the informative modality softly. This paper proposes attentional audio-visual multi-layer feature fusion model, in which soft threshold attention unit are applied on feature mapping at every layer of decoder. The proposed model demonstrates the superior performance of the network against the state-of-the-art models.

eess.AS↗

arXiv2022

GLD-Net: Improving Monaural Speech Enhancement by Learning Global and Local Dependency Features with GLD Block

For monaural speech enhancement, contextual information is important for accurate speech estimation. However, commonly used convolution neural networks (CNNs) are weak in capturing temporal contexts since they only build blocks that process one local neighborhood at a time. To address this problem, we learn from human auditory perception to introduce a two-stage trainable reasoning mechanism, referred as global-local dependency (GLD) block. GLD blocks capture long-term dependency of time-frequency bins both in global level and local level from the noisy spectrogram to help detecting correlations among speech part, noise part, and whole noisy input. What is more, we conduct a monaural speech enhancement network called GLD-Net, which adopts encoder-decoder architecture and consists of speech object branch, interference branch, and global noisy branch. The extracted speech feature at global-level and local-level are efficiently reasoned and aggregated in each of the branches. We compare the proposed GLD-Net with existing state-of-art methods on WSJ0 and DEMAND dataset. The results show that GLD-Net outperforms the state-of-the-art methods in terms of PESQ and STOI.

eess.AS↗

arXiv2022

Multi-layer Feature Fusion Convolution Network for Audio-visual Speech Enhancement

Speech enhancement can potentially benefit from the visual information from the target speaker, such as lip movement and facial expressions, because the visual aspect of speech is essentially unaffected by acoustic environment. In this paper, we address the problem of enhancing corrupted speech signal from videos by using audio-visual (AV) neural processing. Most of recent AV speech enhancement approaches separately process the acoustic and visual features and fuse them via a simple concatenation operation. Although this strategy is convenient and easy to implement, it comes with two major drawbacks: 1) evidence in speech perception suggests that in humans the AV integration occurs at a very early stage, in contrast to previous models that process the two modalities separately at early stage and combine them only at a later stage, thus making the system less robust, and 2) a simple concatenation does not allow to control how the information from the acoustic and the visual modalities is treated. To overcome these drawbacks, we propose a multi-layer feature fusion convolution network (MFFCN), which separately process acoustic and visual modalities for preserving each modality features while fusing both modalities' features layer by layer in encoding phase for enjoying the human AV speech perception. In addition, considering the balance between the two modalities, we design channel and spectral attention mechanisms to provide additional flexibility in dealing with different types of information expanding the representational ability of the convolution neural network. Experimental results show that the proposed MFFCN demonstrates the performance of the network superior to the state-of-the-art models.

eess.AS↗