AMFFCN: Attentional Multi-layer Feature Fusion Convolution Network for Audio-visual Speech Enhancement
This addresses speech enhancement for isolating desired speakers, but it is incremental as it builds on existing encoder-decoder architectures with added attention and fusion mechanisms.
The paper tackles audio-visual speech enhancement by proposing a model that fuses audio and visual features layer-by-layer and uses soft threshold attention to select informative modalities, achieving superior performance against state-of-the-art models.
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.