YGYongjie Guan

Yongjie Guan

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Yongjie Guan has a total of 23 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    ProxLink: A Lightweight Decentralized Synchronization Framework for Multi-User XR

    by Yongjie Guan, Xueyu Hou, Bo Han, Yang Deng, Mingrui Yin, Tao Han

  • Open Access

    MetaGPT: An Immersive Mixed Reality AI Assistant via Offline Edge-Deployed Multi-Agent Vision-Language Inference

    by Mingrui Yin, Wantong Lyu, Sohom Sen, Yongjie Guan, Xueyu Hou, Tao Han, Nirwan Ansari

  • Open Access

    MR-VLMCoach: Immersive Sports Coaching with Vision-Language-Driven Biomechanical Feedback

    by Sohom Sen, Mingrui Yin, Wantong Lyu, Devesh Tiwari, Yongjie Guan, Xueyu Hou, Tao Han, Nirwan Ansari

    Real-time mixed-reality (MR) coaching can make sports feedback more intuitive by showing expert and learner motion in the same visual space. However, most MR coaching systems rely on geometric similarity scores that indicate motion mismatch but do not explain the main error or corrective action. We present MR-VLMCoach, an edge-assisted MR coaching system that combines semantic-aware transmission of motion-critical joints, SMPL-X-based avatar reconstruction, LSTM temporal alignment, sport-adaptive motion-quality scoring, and asynchronous semantic feedback. The system converts ranked joint-level deviations into structured Observation/Impact/Action coaching cues and renders them through avatar overlays, joint highlights, score tiers, text, and synchronized audio. A 100-sample trace study shows that the geometric MR loop remains within the sub-100,ms interaction budget, achieving approximately 79,ms latency under dominant-priority joint transmission, while semantic feedback runs asynchronously off the critical path. Representative case studies across tennis, soccer, basketball, and baseball motions illustrate the cross-sport applicability of the design.

  • Open Access

    LocRender: Local-Depth-Aware Edge Rendering for Responsive Mixed Reality

    by Yongjie Guan, Xueyu Hou

    Mixed reality headsets are expected to support high-fidelity 3D content, but their limited graphics capability, power budget, and thermal capacity make fully local rendering difficult. Edge-assisted rendering can offload expensive graphics computation to a nearby server, but remotely rendered frames may become stale when they arrive at the headset, causing viewpoint mismatch and geometric inconsistency. This paper presents LocRender, a local-depth-aware edge rendering framework for responsive mixed reality. LocRender treats an edge-rendered RGB-D frame as an intermediate representation rather than a final display image. The client uses the latest headset pose to reproject the received frame toward the current viewpoint and uses local depth to identify pixels that conflict with current local geometry. This design allows the headset to act as an active correction endpoint while keeping expensive rendering on the edge side. We implement a controlled RGB-D prototype of LocRender and evaluate it under motion-only and local-occlusion settings. Preliminary results show that RGB-D reprojection improves visual quality under pose staleness, while local-depth-aware correction further improves consistency when the client observes geometry absent from the edge-rendered frame. These results suggest that local sensing can serve as a lightweight correction signal for responsive edge-assisted mixed reality.

  • Open Access

    ImmersiveSlicing: An O-RAN Cross-Layer Reinforcement Learning Framework for Low-Latency Immersive Applications

    by Mingrui Yin, Sohom Sen, Zhihao Ren, Xiaoyu Fang, Yongjie Guan, Tao Han, Nirwan Ansari


  • Falcon: A Mobile Semantic Visual Perception Framework Enlightened by Human Vision Systems

    by Xueyu Hou, Yongjie Guan, Tao Han

    In the realm of computer human visual perception, semantic perception means recognizing objects, people, and scenes. This involves not just detecting shapes and colors but understanding what those visual elements represent. Despite its vital role in enhancing various aspects of emerging applications such as safety for autonomous driving and immersion for mixed reality (MR), real-time segmentation on mobile and edge platforms is challenging due to the nature of dense pixel labeling. To address this issue, we propose $\sf {Falcon}$, a lightweight focus-aware segmentation framework that effectively integrates multiple innovations to achieve real-time segmentation on resource-constrained mobile and edge devices. We design a novel, low-dimension feature for efficient pixel labeling with shallow neural networks, an agile focus-aware refining scheme to compensate for the coarse nature of holistic segmentation, and a modularized design to accommodate the diversity of mobile and edge platforms and ensure seamless integration with different segmentation models. We build a prototype implementation of $\sf {Falcon}$ that supports both on-device executions and edge-assisted offloading, and asynchronous segmentation and refinement. We extensively evaluate the performance of $\sf {Falcon}$ for autonomous driving and MR applications with real setups and standard datasets. Our results demonstrate that $\sf {Falcon}$ achieves real-time segmentation, with an impressive rate of up to 40 frames per second.

  • MRCoach: A Real-Time IoT-Enabled Mixed Reality System With Semantic-Aware Transmission for Smart Sports and Personalized Coaching

    by Mingrui Yin, Sohom Sen, Yongjie Guan, Dhananjay Jagdish Dubey, Xinran Chen, Xueyu Hou, Tao Han, Nirwan Ansari

    Real-time transmission of large-scale data, high computational demands, and resource limitations on edge devices poses significant challenges for intelligent sports systems. The proliferation of the Internet of Things (IoT) technologies has catalyzed the rise of smart sport, where wearable sensors, cameras, and intelligent algorithms are integrated to revolutionize athletic training. Despite this transformation, access to professional coaching remains constrained by factors such as time, cost, and scalability. A critical limitation of existing remote coaching approaches is their inability to perform effective spatiotemporal analysis, hindering comprehensive evaluation and refinement of athletic performance. This article introduces mixed reality coaching (MRCoach), a mixed reality-based, immersive, and interactive sports coaching system that enables data-driven training without requiring in-person supervision. MRCoach reconstructs 3-D volumetric avatars of both learners and expert athletes, allowing users to visualize and compare their movements side-by-side in a mixed reality environment for intuitive skill refinement. To ensure responsive and efficient feedback, we propose an adaptive semantic transmission strategy that prioritizes sport-relevant joints, thereby reducing latency and bandwidth requirements without sacrificing accuracy. Furthermore, a 3-D sports analysis framework is developed to evaluate motion based on normalized joint positions, velocities, and accelerations. This framework computes real-time similarity scores and delivers actionable guidance to learners. Experimental results across four sports—tennis, soccer, basketball, and baseball—demonstrate MRCoach’s effectiveness in providing personalized, real-time training experiences. Compared with state-of-the-art baselines such as MagicStream, ExPose, and PIXIE, MRCoach achieves significantly lower end-to-end (E2E) latency (79.9 ms) and higher frame rates ( $\geq 56$ frames/s) while maintaining accurate pose tracking and high avatar fidelity.


  • Open Access

    ViEdge: Video Analytics on Distributed Edge

    by Xueyu Hou, Yongjie Guan, Tao Han

    With the development of edge computing and increasing demand on video analytics, it is attractive to implement distributed video analytics across edge devices. In this article, we propose ViEdge, a distributed video analytics system across edge devices. ViEdge differs from status quo edge-side video analytics systems in: First, ViEdge does not assume the existence of an edge/cloud server. Instead of processing in a cascaded way between edge devices and server, the video analytics in ViEdge is processed across distributed edge devices in a parallel way. Second, ViEdge addresses two practical challenges in video analytics systems. Specifically, ViEdge optimizes the performance of glance-and-focus object detection pipeline and query related processing with multiple query types. The characters of edge devices (computing capabilities and network conditions) and features of query types (computational complexities and input/output sizes) are comprehensively considered in development of components in ViEdge. By modeling the challenges as multiway number partitioning problems, ViEdge provides practical solutions to optimizing the object detection pipeline and allocations of multiple queries of different types across distributed edge devices. Compared to baseline methods in distributed video analytics across edge devices, ViEdge reaches 1.4 × to 5.3 × speedup in different network environments with neglectable overhead.