Author name cluster

Yujia Wang

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

11 papers

2 author rows

TMLR Journal 2026 Journal Article

ParaBlock: Communication-Computation Parallel Block Coordinate Federated Learning for Large Language Models

Yujia Wang
Yuanpu Cao
Jinghui Chen

Federated learning (FL) has been extensively studied as a privacy-preserving training paradigm. Recently, federated block coordinate descent scheme has become a popular option in training large-scale models, as it allows clients to train only a subset of the model locally instead of the entire model. However, in the era of large language models (LLMs), even a single block can contain a significant number of parameters, posing substantial communication latency, particularly for resource-constrained clients. To address this challenge in federated training/fine-tuning LLMs, we propose ParaBlock, a novel approach that establishes two parallel threads for communication and computation to enhance communication efficiency. We theoretically prove that the proposed ParaBlock achieves the same convergence rate as the standard federated block coordinate descent methods. Empirical evaluations on fine-tuning LLMs on general instruction following and mathematical reasoning confirm that ParaBlock not only maintains strong performance but also significantly improves communication efficiency.

PDF Details

NeurIPS Conference 2025 Conference Paper

AltLoRA: Towards Better Gradient Approximation in Low-Rank Adaptation with Alternating Projections

Xin Yu
Yujia Wang
Jinghui Chen
Lingzhou Xue

Low-Rank Adaptation (LoRA) has emerged as an effective technique for reducing memory overhead in fine-tuning large language models. However, it often suffers from sub-optimal performance compared with full fine-tuning since the update is constrained in the low-rank space. Recent variants such as LoRA-Pro attempt to mitigate this by adjusting the gradients of the low-rank matrices to approximate the full gradient. However, LoRA-Pro's solution is not unique, and different solutions can lead to significantly varying performance in ablation studies. Besides, to incorporate momentum or adaptive optimization design, approaches like LoRA-Pro must first compute the equivalent gradient, causing a higher memory cost close to full fine-tuning. A key challenge remains in integrating momentum properly into the low-rank space with lower memory cost. In this work, we propose AltLoRA, an alternating projection method that avoids the difficulties in gradient approximation brought by the joint update design, meanwhile integrating momentum without higher memory complexity. Our theoretical analysis provides convergence guarantees and further shows that AltLoRA enables stable feature learning and robustness to transformation invariance. Extensive experiments across multiple tasks demonstrate that AltLoRA outperforms LoRA and its variants, narrowing the gap toward full fine-tuning while preserving superior memory efficiency.

PDF Details

AAAI Conference 2025 Conference Paper

Target Scanpath-Guided 360-Degree Image Enhancement

Yujia Wang
Fang-Lue Zhang
Neil A. Dodgson

360° images have wide applications in fields such as virtual reality and user experience design. Our goal is to adjust these images to guide users' visual attention. To achieve this, we present a novel task: target scanpath-guided 360° image enhancement, which aims to enhance 360° images based on user-specified target scanpaths. We develop a Progressive Scanpath-Guided Enhancement Method (PSEM) to address this problem through three stages. In the first stage, we propose a Time-Alignment and Spatial Similarity Clustering (TASSC) algorithm that accounts for the spherical nature of 360° images and the temporal dependency of scanpaths to generate representative scanpaths. In the second stage, we learn the differences between the source and the target scanpaths and select the objects to be edited based on these differences. Particularly, we propose a Dual-Stream Scanpath Difference Encoder (DSDE) embedded into the Segment Anything Model (SAM) network for object mask generation. Finally, we employ a Stable Diffusion network fine-tuned with LoRA technology to produce the final enhanced image. Additionally, we design special loss functions to supervise the training of the second and third stages. Experimental results have demonstrated the effectiveness of our approach for scanpath-guided 360° image enhancement.

PDF Details DOI

EAAI Journal 2024 Journal Article

A depth map stitching framework based on salient region matching

Zetian Mi
Haixia Qi
Jiaxin Chen
Yang Yu
Yujia Wang
Huibing Wang
Xianping Fu

Depth map stitching technology is urgently needed in varying fields such as panoramic three-dimensional reconstruction, underwater mapping, autonomous driving, and robot collision avoidance control. Existing image stitching research mainly focuses on three-primary-color images and heavily relies on feature detection quality, while depth map stitching is rarely studied due to the sparse features and low resolution. To address the above challenges, this paper proposes a color image-guided depth map stitching framework consisting of two stages: coarse depth map stitching and depth correction. In the first stage, a color-image-guided coarse depth map stitching network is proposed to accurately estimate the homography matrix, which can further eliminate parallax artifacts in the stitched depth map. In the second stage, a transformer-based depth correction network is designed, which combines saliency detection and regional matching strategies for depth correction, in order to solve the inconsistency in depth value introduced by point-to-point matching and improve the speed of depth correction to some extent. Extensive experiments demonstrate that the proposed method effectively solves the problem of disparity artifacts and mismatched depth values of the same target, and has excellent performance in depth map stitching.

Details DOI

ICML Conference 2024 Conference Paper

FADAS: Towards Federated Adaptive Asynchronous Optimization

Yujia Wang
Shiqiang Wang
Songtao Lu
Jinghui Chen

Federated learning (FL) has emerged as a widely adopted training paradigm for privacy-preserving machine learning. While the SGD-based FL algorithms have demonstrated considerable success in the past, there is a growing trend towards adopting adaptive federated optimization methods, particularly for the training of large-scale models. However, the conventional synchronous aggregation design poses a significant challenge to the practical deployment of those adaptive federated optimization methods, particularly in the presence of straggler clients. To fill this research gap, this paper introduces federated adaptive asynchronous optimization, named FADAS, a novel method that incorporates asynchronous updates into adaptive federated optimization with provable guarantees. To further enhance the efficiency and resilience of our proposed method in scenarios with significant asynchronous delays, we also extend FADAS with a delay-adaptive learning adjustment strategy. We rigorously establish the convergence rate of the proposed algorithms and empirical results demonstrate the superior performance of FADAS over other asynchronous FL baselines.

Details

TMLR Journal 2024 Journal Article

On the Data Heterogeneity in Adaptive Federated Learning

Yujia Wang
Jinghui Chen

Adaptive federated learning, which benefits from the characteristic of both adaptive optimizer and federated training paradigm, has recently gained lots of attention. Despite achieving outstanding performances on tasks with heavy-tail stochastic gradient noise distributions, adaptive federated learning also suffers from the same data heterogeneity issue as standard federated learning: heterogeneous data distribution across the clients can largely deteriorate the convergence of adaptive federated learning. In this paper, we propose a novel adaptive federated learning framework with local gossip averaging to address this issue. Particularly, we introduce a client re-sampling mechanism and peer-to-peer gossip communications between local clients to mitigate the data heterogeneity without requiring additional gradient computation costs. We theoretically prove the fast convergence for our proposed method under non-convex stochastic settings and empirically demonstrate its superior performances over vanilla adaptive federated learning with client sampling. Moreover, we extend our framework to a communication-efficient variant, in which clients are divided into disjoint clusters determined by their connectivity or communication capabilities. We exclusively perform local gossip averaging within these clusters, leading to an enhancement in network communication efficiency for our proposed method.

PDF Details

AAAI Conference 2024 Conference Paper

Predicting Real-World Penny Auction Durations by Integrating Game Theory and Machine Learning

Yujia Wang
Haoran Yu

Game theory and machine learning are two widely used techniques for predicting the outcomes of strategic interactions among humans. However, the game theory-based approach often relies on strong rationality and informational assumptions, while the machine learning-based approach typically requires the testing data to come from the same distribution as the training data. Our work studies how to integrate the two techniques to address these weaknesses. We focus on the interactions among real bidders in penny auctions, and develop a three-stage framework to predict the distributions of auction durations, which indicate the numbers of bids and auctioneer revenues. Specifically, we first leverage a pre-trained neural network to encode the descriptions of products in auctions into embeddings. Second, we apply game theory models to make preliminary predictions of auction durations. In particular, we tackle the challenge of accurately inferring parameters in game theory models. Third, we develop a Multi-Branch Mixture Density Network to learn the mapping from product embeddings and game-theoretic predictions to the distributions of actual auction durations. Experiments on real-world penny auction data demonstrate that our framework outperforms both game theory-based and machine learning-based prediction approaches.

PDF Details DOI

ICLR Conference 2024 Conference Paper

Tackling the Data Heterogeneity in Asynchronous Federated Learning with Cached Update Calibration

Yujia Wang
Yuanpu Cao
Jingcheng Wu
Ruoyu Chen
Jinghui Chen

Asynchronous federated learning, which enables local clients to send their model update asynchronously to the server without waiting for others, has recently emerged for its improved efficiency and scalability over traditional synchronized federated learning. In this paper, we study how the asynchronous delay affects the convergence of asynchronous federated learning under non-i.i.d. distributed data across clients. Through the theoretical convergence analysis of one representative asynchronous federated learning algorithm under standard nonconvex stochastic settings, we show that the asynchronous delay can largely slow down the convergence, especially with high data heterogeneity. To further improve the convergence of asynchronous federated learning under heterogeneous data distributions, we propose a novel asynchronous federated learning method with a cached update calibration. Specifically, we let the server cache the latest update for each client and reuse these variables for calibrating the global update at each round. We theoretically prove the convergence acceleration for our proposed method under nonconvex stochastic settings. Extensive experiments on several vision and language tasks demonstrate our superior performances compared to other asynchronous federated learning baselines.

Details

ICML Conference 2022 Conference Paper

Communication-Efficient Adaptive Federated Learning

Yujia Wang
Lu Lin 0001
Jinghui Chen

Federated learning is a machine learning training paradigm that enables clients to jointly train models without sharing their own localized data. However, the implementation of federated learning in practice still faces numerous challenges, such as the large communication overhead due to the repetitive server-client synchronization and the lack of adaptivity by SGD-based model updates. Despite that various methods have been proposed for reducing the communication cost by gradient compression or quantization, and the federated versions of adaptive optimizers such as FedAdam are proposed to add more adaptivity, the current federated learning framework still cannot solve the aforementioned challenges all at once. In this paper, we propose a novel communication-efficient adaptive federated learning method (FedCAMS) with theoretical convergence guarantees. We show that in the nonconvex stochastic optimization setting, our proposed FedCAMS achieves the same convergence rate of $O(\frac{1}{\sqrt{TKm}})$ as its non-compressed counterparts. Extensive experiments on various benchmarks verify our theoretical analysis.

Details

IROS Conference 2021 Conference Paper

PNS: Population-Guided Novelty Search for Reinforcement Learning in Hard Exploration Environments

Qihao Liu
Yujia Wang
Xiaofeng Liu

Reinforcement Learning (RL) has made remarkable achievements, but it still suffers from inadequate exploration strategies, sparse reward signals, and deceptive reward functions. To alleviate these problems, a Population-guided Novelty Search (PNS) parallel learning method is proposed in this paper. In PNS, the population is divided into multiple sub-populations, each of which has one chief agent and several exploring agents. The chief agent evaluates the policies learned by exploring agents and shares the optimal policy with all sub-populations. The exploring agents learn their policies in collaboration with the guidance of the optimal policy and, simultaneously, upload their policies to the chief agent. To balance exploration and exploitation, the Novelty Search (NS) is employed in every chief agent to encourage policies with high novelty while maximizing per-episode performance. We apply PNS to the twin delayed deep deterministic (TD3) policy gradient algorithm. The effectiveness of PNS to promote exploration and improve performance in continuous control domains is demonstrated in the experimental section. Notably, PNS-TD3 achieves rewards that far exceed the SOTA methods in environments with sparse or delayed reward signals. We also demonstrate that PNS enables robotic agents to learn control policies directly from pixels for sparse-reward manipulation in both simulated and real-world settings.

Details

AAAI Conference 2019 Conference Paper

3D Face Synthesis Driven by Personality Impression

Yining Lang
Wei Liang
Yujia Wang
Lap-Fai Yu

Synthesizing 3D faces that give certain personality impressions is commonly needed in computer games, animations, and virtual world applications for producing realistic virtual characters. In this paper, we propose a novel approach to synthesize 3D faces based on personality impression for creating virtual characters. Our approach consists of two major steps. In the first step, we train classifiers using deep convolutional neural networks on a dataset of images with personality impression annotations, which are capable of predicting the personality impression of a face. In the second step, given a 3D face and a desired personality impression type as user inputs, our approach optimizes the facial details against the trained classifiers, so as to synthesize a face which gives the desired personality impression. We demonstrate our approach for synthesizing 3D faces giving desired personality impressions on a variety of 3D face models. Perceptual studies show that the perceived personality impressions of the synthesized faces agree with the target personality impressions specified for synthesizing the faces.

PDF Details