Biography

Links

I am now a Ph.D. candidate in the department of Electrical and Computer Engineering at University of California San Diego, supervised by Prof. Florian Meyer. I am also working closely with Prof. Nikolay A. Atanasov and Prof. Mohammad J. Khojasteh. I received my M.S. degree of Computer Engineering at University of California San Diego in 2020 and my B.Eng. degree of Information Engineering at Shanghai Jiao Tong University in 2017.

University of California San Diego

Florian Meyer

Shanghai Jiao Tong University

Mohammad J. Khojasteh

Nikolay A. Atanasov

My research focuses on algorithmic research of

Statistical estimation and signal processing: Bayesian graphic inference, importance sampling, particle filtering.
Information seeking and control: model-based RL, multiagent control, exploration and planning.

The applicable scenario I am solving includes multiobject tracking, SLAM, etc., which are associated with industry like robotics, autonomous driving, wireless communication and smart network, underwater surveillance and localization.

Research Projects

3D Multiobject Tracking with Improved Sampling Efficiency
Considering the multimodality of state density in multisensory environment and particle degeneracy of traditional particle filter, we propose a graphical inference algorithm with beliefs approximately by particle flow.

Selected publications:

[1] W. Zhang and F. Meyer, "Multisensor multiobject tracking with improved sampling efficiency," IEEE Trans. Signal Process., vol. 72, pp. 2036–2053, 2024.

[2] W. Zhang and F. Meyer, "Graph-based multiobject tracking with embedded particle flow," in Proc. IEEE RadarConf-21, Atlanta, GA, USA, May 2021, pp. 1–6.

"Multisensor multiobject tracking with improved sampling efficiency,"

"Graph-based multiobject tracking with embedded particle flow,"

3D Source Localization with Stochastic Particle Flow
In this work, we consider a complex data-associated multiple source localization in a 3D space. The stochastic flow is optimized for stiffness relaxation by solving a stochastic partial differential equation.

Selected publications:

[1] W. Zhang, M. J. Khojasteh, and F. Meyer, "Particle flows for source localization in 3-D using TDOA measurements," in Proc. FUSION-24, Best Student Paper Award, Venice, Italy, Jul. 2024.

[2] W. Zhang, M. J. Khojasteh, N. A. Atanasov, and F. Meyer, "Importance sampling with stochastic particle flow and diffusion optimization," 2024. arXiv: 2412.09778 [eess.SP]

"Particle flows for source localization in 3-D using TDOA measurements,"

"Importance sampling with stochastic particle flow and diffusion optimization,"

Data Fusion and Path Planning for SLAM
We focus on data fusion of estimation layer and path planning of control layer in a SLAM task for machine perception. It's interesting to show how these two layers affect with each other and how information-based seeking can guide the control for improved estimation accuracy.

Selected publications:

[1] W. Zhang, B. Teague, and F. Meyer, "Active planning for cooperative localization: A Fisher information approach," in Proc. Asilomar-22, Pacific Grove, CA, USA, Oct. 2022, pp. 795–800.

[2] E. Leitinger, B. Teague, W. Zhang, M. Liang, and F. Meyer, "Data fusion for radio frequency SLAM with robust sampling," in Proc. FUSION-22, Linkoping, Sweden, 2022, pp. 1–6.

"Active planning for cooperative localization: A Fisher information approach,"

"Data fusion for radio frequency SLAM with robust sampling,"

News

[ 07/2024 ] I presented my work about 3D source localization with particle flow at ISIF/IEEE FUSION 2024, Venice, Italy, and honored the best student paper award.
[ 02/2024 ] My work about 3D MOT with particle flow is accepted to IEEE Transactions on Signal Processing.
[ 06/2023 ] I attended the SIAM MPI workshop about "Bayesian particle flow against stiffness" hold by Dr. Fred Daum from Raytheon Technologies, New Jersey Institute of Technology, Newark, NJ, USA.
[ 10/2022 ] I presented my work about planning for cooperative localization at Asilomar 2022, Pacific Grove, CA, USA.
[ 08/2022 ] I am awarded the Qualcomm Innovation Fellowship, 2022, North America.
[ 05/2021 ] I presented my work about 3D MOT with particle flow at 2021 IEEE Radar Conference online.
[ 09/2020 ] I joined the lab of Prof. Florian Meyer with full-time scholarship as a Ph.D. student.
[ 12/2019 ] I attended and presented work about the Open Speech Platform at 178th Meeting of the Acoustical Society of America, The Hotel del Coronade, San Diego, CA, USA.
[ 10/2018 ] I got an offer as graduate student research (GSR) with Dr. Harinath Garudadri.
[ 07/2018 ] I got an offer for summer research internship program (SRIP) at UCSD, working with Dr. Eric Granholm and Dr. Harinath Garudadri.

Qualcomm Innovation Fellowship

the best student paper award

Biography

Research Projects

3D Multiobject Tracking with Improved Sampling Efficiency Considering the multimodality of state density in multisensory environment and particle degeneracy of traditional particle filter, we propose a graphical inference algorithm with beliefs approximately by particle flow.

3D Source Localization with Stochastic Particle Flow In this work, we consider a complex data-associated multiple source localization in a 3D space. The stochastic flow is optimized for stiffness relaxation by solving a stochastic partial differential equation.

News

3D Multiobject Tracking with Improved Sampling Efficiency
Considering the multimodality of state density in multisensory environment and particle degeneracy of traditional particle filter, we propose a graphical inference algorithm with beliefs approximately by particle flow.

3D Source Localization with Stochastic Particle Flow
In this work, we consider a complex data-associated multiple source localization in a 3D space. The stochastic flow is optimized for stiffness relaxation by solving a stochastic partial differential equation.