Lab. Intro.

Pervasive Perception and Computing Systems (PPCS) Lab @ National Taipei University of Technology

The PPCS Lab is currently being established at National Taipei University of Technology. We welcome motivated M.S. students and undergraduate researchers interested in pervasive intelligence, edge AI, multimodal intelligent sensing, AIoT, and cyber-physical systems.

Research Directions

  • Pervasive intelligence
  • Edge AI
  • Multimodal intelligent sensing
  • AIoT
  • Cyber-physical systems

Selected Thesis Topics from Previous Supervision

The following thesis topics are selected examples from previous student supervision and research mentoring, including work conducted in the Mobile Sensing and Systems (MSS) Lab at Yuan Ze University. They are organized by research area to illustrate the evolving directions and project scope related to the lab.

Federated Learning and Edge Intelligence
  • 2025 — Pseudo-Label Guided Federated Learning for Handling Missing Modalities (Bo-Sheng Liao, M.S., YZU, in-service)
  • 2025 — A Study of Imputation Methods for Missing Values in Multi-Modal Federated Learning (Chin Fan-Chiang, M.S., YZU)
  • 2025 — Deep Reinforcement Learning Based Model Selections on an Edge Device Considering Latency and Accuracy (Yi-Chun Lin, M.S., YZU)
  • 2024 — Client Attributes-aware Selection using Reinforcement Learning in Federated Learning (Yi-Chien Wu, M.S., YZU)
  • 2023 — Learning-based Client Selections Considering Client Similarity in Federated Learning (Hui-Rou Zhoug, M.S., YZU)
  • 2023 — Channel Pruning with Hybrid Approaches for Convolutional Neural Networks (Argha Orion Silitonga, M.S., YZU)
  • 2022 — Dynamic Clustering Control for Communication-efficient Federated Learning (Guan-Lin Chen, M.S., YZU)
Intelligent Transportation and Vehicular Systems
  • 2025 — Leveraging Large Language Models for Multi-Intersection Traffic Signal Control (Cheng-Ying Chen, M.S., YZU)
  • 2024 — A See-through System Using Deep Homography in Vehicular Networks (Ming-Wei Yi, M.S., YZU)
  • 2020 — Missing Traffic Data Imputation Considering Multi-System Data using Generative Adversarial Nets (Kuan-Ju Huang, M.S., YZU)
  • 2020 — Traffic Flow Predictions of Transportation Systems using Multi-view and Multi-task Learning (Ting-Wei Fan, M.S., YZU)
  • 2018 — A Lightweight Augmented Reality System to See-through Car (Bing-Jie Yuan, M.S., YZU)
  • 2018 — A Critical-spot Aware Approach for Prediction of Hard Brake Maneuver (Tai-Lin Kuo, M.S., YZU)
Human-centered Computing and Social Intelligence
  • 2024 — Research on Social-aware Robot in an Indoor Environment (Sevendi Eldrige Rifki Poluan, Ph.D., YZU)
  • 2022 — Robot Trajectory Generation for Approaching a Free-standing Conversational Group Considering Human Body Skeleton (Min-Hung Cheng, M.S., YZU)
  • 2021 — An Interaction-aware CNN-LSTM Model for Predicting Romantic Interests (Marvela Victorya Rampengan, M.S., YZU)
  • 2019 — Audio-Visual Fusion for Conversation Group Detection (Chang-Hua Chuang, M.S., YZU)
  • 2019 — Using Smart Insoles and RGB Camera for Identifying Stationary Human Targets (Sevendi Eldrige Rifki Poluan, M.S., YZU)
Multimodal Sensing and Intelligent Systems
  • 2024 — Enhancing Ship Recognition in Infrared Images through Style-Transformed Image Augmentation (Hung-Wen Lin, M.S., YZU, in-service)
  • 2023 — Adaptive Annotations of Density Estimation using Deep Learning in Dense Object Counting (An-An Li, M.S., YZU)
  • 2023 — Design and Implement of an AI-powered Desktop Coffee Bean Selection System (Jai-Wei Wu, M.S., YZU, in-service)
  • 2023 — Acoustic-based Placement Recognition of Smart Device using Deep Learning (Jian-Cheng Gan, M.S., YZU, in-service)
  • 2021 — Audiovisual Crowd Counting based on Image Inpainting of Crowd Density Map under Complex Scenes (Chun-Yu Kuo, M.S., YZU)
  • 2021 — Design and Implementation of a Gesture Recognition System using Light Sensors (Tsung-Yen Ho, M.S., YZU, in-service)
  • 2021 — Research on the Performance Evaluation of Wireless Network Protocols under TaiSEIA 101 (Shao-Yi Chen, M.S., YZU, in-service)

歷年大學部專題題目

以下整理歷年指導之大學部專題題目，主題涵蓋 AIoT、邊緣智慧、電腦視覺、智慧交通、機器人與無人系統、人本感測與應用型人工智慧系統。

AIoT、邊緣智慧與智慧感知
  • 邊緣裝置上應用光感測於手勢辨識系統（2022）
  • 基於YOLOv5與模型減量建立即時服務人員辨識與盲人引導系統（2022）
  • FedSAUC: A Similarity-Aware Update Control For Communication-Efficient Federated Learning in Edge Computing（2021）
  • 基於聲音資訊之賣場銷售員銷售統計系統（2019）
  • 飲水監測系統（2017）
電腦視覺與智慧交通
  • 針對交通情境之開放詞彙多模態異常偵測與輕量化分析（2026）
  • Zero-Shot Vehicle Model Recognition via Text-Based Retrieval-Augmented Generation（2025）
  • 應用於交通異常偵測之輕量化影像語言模型（2024）
  • 在車聯網架構下基於神經網路生成車透視影像（2023）
  • Recognizing Driver Talking Direction to Increase Driving Safety Based on Deep Learning（2021）
  • 車輛駕駛輔助系統（2018）
機器人與無人系統
  • 透過 VLM 開發具有主持撲克遊戲能力的社交機器人（2025）
  • 基於單攝影機之小型無人機於室內軌跡自動重複方法（2024）
  • 具身分辨識隱私性之商品導覽機器人（2022）
  • 具室內場景辨識之機器人系統（2020）
  • 你丟我撿: 具意圖偵測之智慧分類環保垃圾桶（2020）
  • 基於影像之害蟲辨識與雷射去除裝置（2018）
人本感測與應用型人工智慧
  • 藉由模態對齊訓練之手語即時辨識系統（2026）
  • 利用穿戴式眼鏡協助過濾青少年接觸之不良資訊（2023）
  • AR與VR裝置利用眼睛瞳孔移動與反應之情緒辨識（2021）
  • 應用生理訊號之情緒辨識於表情符號推薦（2020）
  • 應用神經網路基於政府開放資料之登革熱疫情預測（2020）
  • 運用深度學習之室內使用情境感知（2018）
  • 室內活動偵測系統（2017）
  • 飲水監測系統（2017）