Hansi Liu
Senior Research Engineer, Mobile Platform Solutions
Samsung Research America
Mountain View, CA
About
At Samsung Research America I work on LLM systems for mobile security and agent safety: detecting when an LLM agent's actions diverge from the user's intent, and LLM + knowledge-graph pipelines that map vulnerability reports to affected devices.
Before joining Samsung, I was a Ph.D. Candidate at Rutgers University's Wireless Information Network Lab (WINLAB) advised by Dr. Marco Gruteser and Dr. Kristin Dana. My research interests include mobile computing and computer vision, with a special focus on multi-modal sensing and collaborative perception. My dissertation focuses on fusing wireless communication with vision sensing to improve systems' performance including sensing range, efficiency, tracking and localization.
Selected Research
Vi-Fi: Cross-Modal Association
Cameras see people; wireless networks see their phones. Vi-Fi associates the two, matching each pedestrian detected in video with a smartphone identifier, without facial recognition. A multi-modal deep network learns affinity matrices from camera bounding boxes, WiFi Fine Time Measurements and IMU data. We built a live demo that finds a target phone's owner in surveillance footage, and released the Vi-Fi dataset.
IPSN'22 paper Vi-Fi dataset MobiSys'21 demo paper Demo slides Demo video 1 Demo video 2
FusionEye: Perception Sharing for Connected Vehicles
A vehicle's sensors stop at the nearest occlusion. FusionEye lets vehicles share perception data over V2V links and merge it into a single, more complete traffic scene, using a self-adaptive topology-merging algorithm on a bipartite graph. It reaches 88% merging accuracy at 5% packet loss and ~200 ms latency. A follow-up added deep appearance and kinematic embeddings for vehicle verification, raising accuracy to 92%.
SECON'19 paper SECON'19 slides MobiSys'19 Rising Stars paper Rising Stars slides
WiFi FTM & Localization
Two studies of WiFi Fine Time Measurement (802.11-2016) for ranging. First, an open experimentation platform and measurement framework: meter-level accuracy is achievable, but only consistently in low-multipath settings or at ≥80 MHz bandwidth. Second, Wi-Go, which fuses FTM, GPS and odometry for lane-level vehicle positioning: 1.3 m median error in Manhattan's urban canyons, 0.8 m in suburban tests.
Publications
Patents — 6 granted, US
| Systems and Methods for Matching Objects in Collaborative Perception Messages Using Multiple Adaptive Thresholds Hansi Liu, Hongsheng Lu, Rui Guo · Toyota | US 12,062,282 Aug 13, 2024 |
| Systems and Methods for Matching Objects in Collaborative Perception Messages Hansi Liu, Hongsheng Lu, Rui Guo · Toyota | US 11,765,562 Sep 19, 2023 |
| Sparse Excitation Method for 3-dimensional Underground Cable Localization by Fiber Optic Sensing Ming-Fang Huang, Ting Wang, Hansi Liu · NEC | US 11,681,042 Jun 20, 2023 |
| Underground Optical Fiber Cable Localization Including DFOS and TDOA Methods Ming-Fang Huang, Hansi Liu, Ting Wang · NEC | US 11,619,523 Apr 4, 2023 |
| Method and Apparatus for Enabling Sequential Groundview Image Projection Synthesis and Complicated Scene Reconstruction at Map Anomaly Hotspot Fan Bai, Hansi Liu, David A. Craig · General Motors | US 11,143,514 Oct 12, 2021 |
| Providing Information-rich Map Semantics to Navigation Metric Map Hansi Liu, Fan Bai, Shuqing Zeng · General Motors | US 10,671,860 Jun 2, 2020 |
News
- Our paper "ViFiCon: Vision and Wireless Association Via Self-Supervised Contrastive Learning" has been accepted at the 9th Multimodal Learning and Applications Workshop (MULA 2026), held in conjunction with CVPR 2026.
- US Patent No. 12,062,282, "Systems and Methods for Matching Objects in Collaborative Perception Messages Using Multiple Adaptive Thresholds" (with Hongsheng Lu and Rui Guo, Toyota), has been granted.
- US Patent No. 11,765,562, "Systems and Methods for Matching Objects in Collaborative Perception Messages" (with Hongsheng Lu and Rui Guo, Toyota), has been granted.
- Our paper "ViFiT: Reconstructing Vision Trajectories from IMU and Wi-Fi Fine Time Measurements" has been accepted for publication in MobiCom ISACom 2023.
- Our paper "ViFi-Loc: Multi-modal Pedestrian Localization using GAN with Camera-Phone Correspondences" has been accepted for publication in ICMI 2023.
- US Patent No. 11,681,042, "Sparse Excitation Method for 3-dimensional Underground Cable Localization by Fiber Optic Sensing" (with Ming-Fang Huang and Ting Wang, NEC Labs America), has been granted.
Older news
- US Patent No. 11,619,523, "Underground Optical Fiber Cable Localization Including DFOS and TDOA Methods" (with Ming-Fang Huang and Ting Wang, NEC Labs America), has been granted.
- I have joined Samsung Research America in Mountain View, CA as a Senior Research Engineer in the Mobile Platform Solutions group.
- I defended my Ph.D.
- Our demo "Tagging Vision with Smartphone Identities by Vision2Phone Translation" received the Best Demonstration Award at SECON'22.
- Our paper "ViTag: Online WiFi Fine Time Measurements Aided Vision-Motion Identity Association in Multi-person Environments" is accepted by SECON'22.
- Our paper "Vi-Fi: Associating Moving Subjects across Vision and Wireless Sensors" is accepted by IPSN'22. We released the Vi-Fi Dataset, a large-scale multi-modal dataset comprising vision and wireless data across multiple indoor/outdoor scenarios.
- US Patent No. 11,143,514, "Method and Apparatus for Enabling Sequential Groundview Image Projection Synthesis and Complicated Scene Reconstruction at Map Anomaly Hotspot" (with Fan Bai and David A. Craig, General Motors), has been granted.
- I am excited to join the ITL group at Toyota North America for a summer internship. I will be working with Dr. Hongsheng Lu on topics of Collaborative Perception Messages (CPM) for connected vehicle systems.
- Our paper "Demo: Lost and Found! Associating Target Persons in Camera Surveillance Footage with Smartphone Identifiers" is accepted by the demo session of MobiSys'21.
- Our paper "New Methods for Non-Destructive Underground Fiber Localization using Distributed Fiber Optic Sensing Technology" is accepted for publication at OECC 2020.
- US Patent No. 10,671,860, "Providing Information-rich Map Semantics to Navigation Metric Map" (with Fan Bai and Shuqing Zeng, General Motors), has been granted.
- Our paper "Wi-Go: Accurate and Scalable Vehicle Positioning using WiFi Fine Timing Measurement" is accepted for publication at ACM MobiSys 2020.
- I passed my Ph.D. qualification exam. Thanks to the committee members (Prof. Dipankar Raychaudhuri, Prof. Yingying Chen, Maria Striki and Prof. Kristin Dana) for their valuable comments and feedback!
- I will be joining NEC Labs America in Princeton this summer. I will work with Ting Wang on topics of underground optical fiber sensing and localization.
- Our paper "FusionEye: Perception Sharing for Connected Vehicles and its Bandwidth-Accuracy Trade-offs" is accepted by SECON 2019.
- I will be joining the General Motors R&D Department for a summer internship and work on topics of connected vehicle systems.