Today, many state-of-the-art automatic speech recognition (ASR) systems apply all-neural
models that map audio to word sequences trained end-to-end along one global optimisation …

Language models (LMs) pre-trained on massive amounts of text, in particular bidirectional
encoder representations from Transformers (BERT), generative pre-training (GPT), and GPT-2…

Background Meishan is a pig breed indigenous to China and famous for its high fecundity.
The traits of Meishan are strongly associated with its distinct evolutionary history and …

The Toll signaling pathway was initially identified for its involvement in the control of early
embryogenesis. It was later shown to be also part of a major innate immune pathway …

Self-supervised-learning-based pre-trained models for speech data, such as Wav2Vec 2.0 (W2V2),
have become the backbone of many speech tasks. In this paper, to achieve speaker …

Automatic speech recognition (ASR) of single channel far-field recordings with an unknown
number of speakers is traditionally tackled by cascaded modules. Recent research shows …

Structural variants (SVs) represent essential forms of genetic variation, and they are associated
with various phenotypic traits in a wide range of important livestock species. However, …

DNA methylation changes play essential roles in regulating the activities of genes involved
in immune responses. Understanding of variable DNA methylation linked to immune …

End-to-end automatic speech recognition (ASR) and large language models, such as Whisper
and GPT-2, have recently been scaled to use vast amounts of training data. Despite the …

Background Reproductive performance of livestock is an economically important aspect of
global food production. The Chinese Meishan pig is a prolific breed, with an average of three …