Wenming Xiao

TL;DR: This paper pre-train a sentiment-aware language model (SentiX) via domain-invariant sentiment knowledge from large-scale review datasets, and utilize it for cross-domain sentiment analysis task without fine-tuning.

TL;DR: Zhang et al. as mentioned in this paper proposed an end-to-end vision-language pre-trained model for both V+L understanding and generation, where a unified Transformer framework was built to jointly learn visual representation and semantic alignments between image and text.

TL;DR: In this article , the impact of factors involved in detection of inherited variants with whole genome sequencing (WGS) was dissected for the implementation of precision medicine and is a complicated process in which each step affects variant call quality.

TL;DR: The genetic built-in-truth of the Quartet family design not only improved sensitivity of benchmark calls by removing additional false positive variants with apparently high quality, but also enabled estimation of the precision of variants calls outside the benchmark regions.

TL;DR: NeuSomatic as discussed by the authors uses the first comprehensive and well-characterized somatic reference data sets from the SEQC2 consortium to investigate best practices for using a deep learning framework in cancer mutation detection.