Institution profile

National University of Defense Technology

Academic institutionasia · cn
Official website
Research library877linked papers
Opportunities0open roles
Selected work

Representative Papers

Prompting Depth Anything for 4K Resolution Accurate Metric Depth Estimation

Dec 18, 2024arXiv.org

To resolve the accuracy–cost trade-off between low-cost LiDAR and high-resolution metric depth estimation, this paper introduces Prompt Depth Anything—a novel paradigm that leverages sparse, low-accuracy LiDAR point clouds as multi-scale geometric prompts to guide the Depth Anything foundation model toward 4K-resolution metric depth prediction. Methodologically, we design a lightweight prompt fusion architecture enabling cross-scale feature alignment and develop a scalable data pipeline integrating LiDAR physics simulation with pseudo-ground-truth generation from real-world scenes. Evaluated on ARKitScenes and ScanNet++, our approach achieves state-of-the-art performance, reducing 4K depth error by 21.3% relatively. Moreover, the high-fidelity depth maps substantially enhance downstream applications, including photorealistic 3D reconstruction and general-purpose robotic grasping.

8 citationsRead paper

RoME: Domain-Robust Mixture-of-Experts for MILP Solution Prediction across Domains

Nov 04, 2025

Existing learning-based methods for accelerating mixed-integer linear programming (MILP) solvers suffer from poor generalization and limited cross-domain transferability. To address this, we propose a task-embedding-driven initial solution prediction framework designed for cross-domain generalization. Our method introduces a Mixture-of-Experts architecture with dynamic instance routing and incorporates two-level distributionally robust optimization: inter-domain alignment and intra-domain perturbation-robust training—enhancing model adaptability to unseen problem distributions. To our knowledge, this is the first approach enabling efficient zero-shot transfer of a single model across diverse domains—including combinatorial optimization, scheduling, and network design. Evaluated on five benchmark domains, it achieves an average speedup of 67.7% over standard solvers. Moreover, on challenging real-world instances from MIPLIB, it significantly outperforms state-of-the-art methods, demonstrating both strong generalization capability and practical utility.

6 citationsRead paper

Deep Temporal Graph Clustering: A Comprehensive Benchmark and Datasets

Aug 07, 2025IEEE Transactions on Pattern Analysis and Machine Intelligence

Temporal graph clustering (TGC) has been hindered by the absence of dedicated methodologies and high-quality datasets. This work addresses this gap by systematically constructing BenchTGC, the first comprehensive benchmark for TGC, and introducing a unified framework tailored to the dynamic nature of temporal graphs. The framework integrates interaction sequence batching, temporal graph neural networks, and dynamic graph representation learning to adapt and reformulate existing clustering algorithms for temporal settings. Alongside the framework, we release the first dataset suite specifically designed for TGC. Extensive experiments demonstrate that BenchTGC substantially improves clustering performance, underscoring the necessity and effectiveness of specialized TGC approaches in balancing spatiotemporal modeling demands and capturing real-world dynamic graph patterns.

5 citationsRead paper

MToP: A MATLAB Optimization Platform for Evolutionary Multitasking

Dec 13, 2023arXiv.org

The multi-task optimization (MTO) community lacks a unified, open-source evaluation platform, hindering reproducibility, fair benchmarking, and practical validation of evolutionary multitasking (EMT) algorithms. Method: This paper introduces EMT-Platform—the first open-source MATLAB platform dedicated to EMT research—featuring a modular architecture, plugin-based algorithm interfaces, a graphical user interface, and built-in knowledge transfer mechanisms. It integrates over 50 multitasking algorithms (including systematically adapted single-task baselines), 200+ standardized benchmark problem instances, and 20+ performance metrics. Contribution/Results: EMT-Platform enables standardized algorithm development, rigorous cross-algorithm evaluation, and intuitive result visualization. It significantly enhances reproducibility, accelerates empirical research, and supports real-world application validation across diverse domains. Widely adopted by the EMT research community, it serves as a foundational infrastructure for advancing both theoretical and applied multitasking optimization.

2 citations1 influentialRead paper

A Universal Load Balancing Principle and Its Application to Large Language Model Serving

Jan 25, 2026

This work addresses the challenges of straggler nodes and high idle time in barrier-synchronized, stateful systems—such as large language model (LLM) inference—caused by workload heterogeneity and clock drift. The authors propose a general dynamic load balancing approach that models workload drift as a non-decreasing stochastic process and employs finite-horizon integer optimization to achieve efficient load distribution under the constraints that tasks are non-migratable and progress is bottlenecked by the slowest node. This study establishes the first general theoretical framework for load balancing in barrier-synchronized systems with worst-case performance guarantees. The benefits of the method amplify with increasing batch size and system scale. Experimental results demonstrate significant improvements in throughput, latency, and energy efficiency, reducing per-step decoding idle computation time by over 40% in real-world LLM serving scenarios.

2 citationsRead paper
Recent publications

Latest Papers