Yinda Xu

ORCID: 0000-0003-3910-685X
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Video Surveillance and Tracking Methods
  • Autonomous Vehicle Technology and Safety
  • Privacy-Preserving Technologies in Data
  • Internet Traffic Analysis and Secure E-voting
  • Human Pose and Action Recognition
  • IoT-based Smart Home Systems
  • Infrared Target Detection Methodologies
  • Robotics and Sensor-Based Localization
  • Millimeter-Wave Propagation and Modeling
  • UAV Applications and Optimization
  • Robotics and Automated Systems
  • Air Quality Monitoring and Forecasting
  • Bayesian Modeling and Causal Inference
  • Cryptography and Data Security
  • Video Analysis and Summarization
  • Topic Modeling
  • Human-Automation Interaction and Safety
  • Indoor and Outdoor Localization Technologies
  • Fire Detection and Safety Systems
  • Machine Learning and Data Classification
  • Radio Wave Propagation Studies
  • Vehicular Ad Hoc Networks (VANETs)
  • Mobile Crowdsensing and Crowdsourcing
  • Transportation and Mobility Innovations
  • Advanced Neural Network Applications

Shanghai Jiao Tong University
2024-2025

Shandong Jiaotong University
2025

Aalto University
2024

Tampere University
2023

Zhejiang University
2019-2020

China Electronics Corporation (China)
2019

Visual tracking problem demands to efficiently perform robust classification and accurate target state estimation over a given at the same time. Former methods have proposed various ways of estimation, yet few them took particularity visual itself into consideration. Based on careful analysis, we propose set practical guidelines for high-performance generic object tracker design. Following these guidelines, design our Fully Convolutional Siamese tracker++ (SiamFC++) by introducing both...

10.1609/aaai.v34i07.6944 article EN Proceedings of the AAAI Conference on Artificial Intelligence 2020-04-03

Trained on massive publicly available data, large language models (LLMs) have demonstrated tremendous success across various fields. While more data contributes to better performance, a disconcerting reality is that high-quality public will be exhausted in few years. In this paper, we offer potential next step for contemporary LLMs: collaborative and privacy-preserving LLM training the underutilized distributed private via federated learning (FL), where multiple owners collaboratively train...

10.1145/3637528.3671582 article EN Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining 2024-08-24

The Vision Meets Drone (VisDrone2019) Single Object Tracking challenge is the second annual research activity focusing on evaluating single-object tracking algorithms drones, held in conjunction with International Conference Computer (ICCV 2019). VisDrone-SOT2019 Challenge goes beyond its VisDrone-SOT2018 predecessor by introducing 25 more challenging sequences for long-term tracking. We evaluate and discuss results of 22 participating 19 state-of-the-art trackers collected dataset. are...

10.1109/iccvw.2019.00029 article EN 2019-10-01

Vehicle-to-everything-aided autonomous driving (V2X-AD) has a huge potential to provide safer solution. Despite extensive research in transportation and communication support V2X-AD, the actual utilization of these infrastructures resources enhancing performances remains largely unexplored. This highlights necessity collaborative driving; that is, machine learning approach optimizes information sharing strategy improve performance each vehicle. effort necessitates two key foundations:...

10.1109/tpami.2025.3560327 article EN IEEE Transactions on Pattern Analysis and Machine Intelligence 2025-01-01

Visual tracking problem demands to efficiently perform robust classification and accurate target state estimation over a given at the same time. Former methods have proposed various ways of estimation, yet few them took particularity visual itself into consideration. After careful analysis, we propose set practical guidelines for high-performance generic object tracker design. Following these guidelines, design our Fully Convolutional Siamese tracker++ (SiamFC++) by introducing both...

10.48550/arxiv.1911.06188 preprint EN other-oa arXiv (Cornell University) 2019-01-01

Trained on massive publicly available data, large language models (LLMs) have demonstrated tremendous success across various fields. While more data contributes to better performance, a disconcerting reality is that high-quality public will be exhausted in few years. In this paper, we offer potential next step for contemporary LLMs: collaborative and privacy-preserving LLM training the underutilized distributed private via federated learning (FL), where multiple owners collaboratively train...

10.48550/arxiv.2402.06954 preprint EN arXiv (Cornell University) 2024-02-10

Autonomous driving systems are always built on motion-related modules such as the planner and controller. An accurate robust trajectory tracking method is indispensable for these a primitive routine. Current methods often make strong assumptions about model context dynamics, which not enough to deal with changing scenarios in real-world system. In this paper, we propose Deep Reinforcement Learning (DRL)-based autonomous systems. The representation learning ability of DL exploration nature RL...

10.48550/arxiv.2308.15991 preprint EN other-oa arXiv (Cornell University) 2023-01-01

In this paper, we propose a novel algorithm for relationship detection. This task involves the tracking of target object and human pose. The is tracked with visual tracker. poses are estimated via keypoint detector while person identities preserved simple yet effective IoU Finally, possessing inference made based on position information humans. meets real-time requirement by running at over 20 FPS give an application illustration in sports analytics.

10.23919/ccc50068.2020.9189516 article EN 2020-07-01

Vehicle-to-everything-aided autonomous driving (V2X-AD) has a huge potential to provide safer solution. Despite extensive researches in transportation and communication support V2X-AD, the actual utilization of these infrastructures resources enhancing performances remains largely unexplored. This highlights necessity collaborative driving: machine learning approach that optimizes information sharing strategy improve performance each vehicle. effort necessitates two key foundations: platform...

10.48550/arxiv.2404.09496 preprint EN arXiv (Cornell University) 2024-04-15

With recent advancements in large language models (LLMs), alignment has emerged as an effective technique for keeping LLMs consensus with human intent. Current methods primarily involve direct training through Supervised Fine-tuning (SFT) or Reinforcement Learning from Human Feedback (RLHF), both of which require substantial computational resources and extensive ground truth data. This paper explores efficient method aligning black-box using smaller models, introducing a model-agnostic...

10.48550/arxiv.2405.18718 preprint EN arXiv (Cornell University) 2024-05-28

The rapid growth of chronic diseases and medical conditions (e.g. obesity, depression, diabetes, respiratory musculoskeletal diseases) in many OECD countries has become one the most significant wellbeing problems, which also poses pressure to sustainability healthcare economies. Thus, it is important promote early diagnosis, intervention, healthier lifestyles. One partial solution problem extending long-term health monitoring from hospitals natural living environments. It been shown...

10.23919/date56975.2023.10137048 article EN Design, Automation & Test in Europe Conference & Exhibition (DATE), 2015 2023-04-01
Coming Soon ...