Sorry, you need to enable JavaScript to visit this website.

Multimodal speech emotion recognition (MSER) is an emerging and challenging field of research due to its more robust characteristics than unimodal. However, in multimodal approaches, the interactive relations for model building using different modalities of speech representations for emotion recognition have not been well investigated yet. To address this issue, we introduce a new approach to capturing the deep temporal features of audio and text. The audio features are learned with a convolution neural network (CNN) and a Bi-directional Gated Recurrent Unit (Bi-GRU) network.

Federated learning can be used to train machine learning models on the edge on local data that never leave devices, providing privacy by default. This presents a challenge pertaining to the communication and computation costs associated with clients’ devices. These costs are strongly correlated with the size of the model being trained, and are significant for state-of-the-art automatic speech recognition models.We propose using federated dropout to reduce the size of client models while training a full-size model server-side.

The recently proposed conformer architecture has been successfully used for end-to-end automatic speech recognition (ASR) architectures achieving state-of-the-art performance on different datasets. To our best knowledge, the impact of using conformer acoustic model for hybrid ASR is not investigated. In this paper, we present and evaluate a competitive conformer-based hybrid model training recipe. We study different training aspects and methods to improve worderror-rate as well as to increase training speed.

Transformer-based architectures have been the subject of research aimed at understanding their overparameterization and the non-uniform importance of their layers. Applying these approaches to Automatic Speech Recognition, we demonstrate that the state-of-the-art Conformer models generally have multiple ambient layers. We study the stability of these layers across runs and model sizes, propose that group normalization may be used without disrupting their formation, and examine their correlation with model weight updates in each layer.

This paper describes our RoyalFlush system for the track of multi-speaker automatic speech recognition (ASR) in the M2MeT challenge. We adopted the serialized output training (SOT) based multi-speakers ASR system with large-scale simulation data. Firstly, we investigated a set of front-end methods, including multi-channel weighted predicted error (WPE), beamforming, speech separation, speech enhancement and so on, to process training, validation and test sets. But we only selected WPE and beamforming as our frontend methods according to their experimental results.