Haiyang Yu

TL;DR: An open-source and extensible knowledge extraction toolkit DeepKE, supporting complicated low-resource, document-level and multimodal scenarios in knowledge base population, implements various information extraction tasks, including named entity recognition, relation extraction and attribute extraction.

TL;DR: This work proposes a new technique, named feature momentum (FM), that uses a momentum step to incorporate historical embeddings when updating feature representations, and develops two specific algorithms, known as Graph FM-IB and GraphFM-OB, that consider in-batch and out-of-batch data, respectively.

TL;DR: Results from a large-scale experiment on pretraining encoders with non-embedding parameter counts ranging from 700M to 9.3B, their subsequent distillation into smaller models, and their application to the Natural Language Understanding (NLU) component of a virtual assistant system are presented.

TL;DR: In this paper , a dynamic communication topology-based CAV-following model is proposed to mitigate the impact of communication failures on platoons, and an adaptive Kalman filter (AKF) method is applied to reduce perceptual error.

TL;DR: In this paper , a novel framework called parallel vehicular crowd sensing (P-VCS) is proposed to balance the physical environment, cyber networks, and human and social factors.