Xuefei Li

Houqiang Zhong, Shaocheng Shen, Ke Cai et al.

Fine-tuning large pre-trained vision foundation models in a parameter-efficient manner is critical for downstream vision tasks, considering the practical constraints of computational and storage costs. Low-rank adaptation (LoRA) is a well-established technique in this domain, achieving impressive efficiency by reducing the parameter space to a low-rank form. However, developing more advanced low-rank adaptation methods to reduce parameters and memory requirements remains a significant challenge in resource-constrained application scenarios. In this study, we consider on top of the commonly used vision transformer and propose Serial LoRA, a novel LoRA variant that introduces a shared low-rank matrix serially composite with the attention mechanism. Such a design extracts the underlying commonality of parameters in adaptation, significantly reducing redundancy. Notably, Serial LoRA uses only 1/4 parameters of LoRA but achieves comparable performance in most cases. We conduct extensive experiments on a range of vision foundation models with the transformer structure, and the results confirm consistent superiority of our method.

Kun Guo, Xuefei Li, Xijun Wang et al.

Federated learning (FL) and split learning (SL) are two effective distributed learning paradigms in wireless networks, enabling collaborative model training across mobile devices without sharing raw data. While FL supports low-latency parallel training, it may converge to less accurate model. In contrast, SL achieves higher accuracy through sequential training but suffers from increased delay. To leverage the advantages of both, hybrid split and federated learning (HSFL) allows some devices to operate in FL mode and others in SL mode. This paper aims to accelerate HSFL by addressing three key questions: 1) How does learning mode selection affect overall learning performance? 2) How does it interact with batch size? 3) How can these hyperparameters be jointly optimized alongside communication and computational resources to reduce overall learning delay? We first analyze convergence, revealing the interplay between learning mode and batch size. Next, we formulate a delay minimization problem and propose a two-stage solution: a block coordinate descent method for a relaxed problem to obtain a locally optimal solution, followed by a rounding algorithm to recover integer batch sizes with near-optimal performance. Experimental results demonstrate that our approach significantly accelerates convergence to the target accuracy compared to existing methods.

Huiwei Zhou, Xuefei Li, Weihong Yao et al.

Protein-protein interaction (PPI) extraction from published scientific literature provides additional support for precision medicine efforts. Meanwhile, knowledge bases (KBs) contain huge amounts of structured information of protein entities and their relations, which can be encoded in entity and relation embeddings to help PPI extraction. However, the prior knowledge of protein-protein pairs must be selectively used so that it is suitable for different contexts. This paper proposes a Knowledge Selection Model (KSM) to fuse the selected prior knowledge and context information for PPI extraction. Firstly, two Transformers encode the context sequence of a protein pair according to each protein embedding, respectively. Then, the two outputs are fed to a mutual attention to capture the important context features towards the protein pair. Next, the context features are used to distill the relation embedding by a knowledge selector. Finally, the selected relation embedding and the context features are concatenated for PPI extraction. Experiments on the BioCreative VI PPI dataset show that KSM achieves a new state-of-the-art performance (38.08% F1-score) by adding knowledge selection.