黃敬群

Semantic Knowledge Learning for Remote Sensing Image Understanding: Learning, Representation, and Accumulation of Semantic Knowledge

資訊科學與工程研究所・共同指導：鄭文皇

遙測影像理解、基礎模型、弱監督語意分割、持續學習、Remote Sensing Image Understanding、Foundation Vision–Language Models、Semantic Knowledge Learning、Weakly Supervised Semantic Segmentation、Continual Learning

DAP-MR: Degradation-Aware Prompt-Magnitude Routing for Controllable Real-World Ultrasound Super-Resolution

電機資訊國際碩士學位學程

超解析度、擴散模型、超音波、Super-Resolution、Diffusion、Ultrasound

Explicit Frictional Contact Modeling in Neural Garment Simulation

電機資訊國際碩士學位學程

神經網路服裝模擬、摩擦接觸、庫侖摩擦、布料模擬、圖神經網路、neural garment simulation、frictional contact、Coulomb friction、cloth simulation、graph neural networks

Learning Ordinal Degradation Representations with Textual Priors for Diffusion-Based Blind Image Super-Resolution

多媒體工程研究所

盲影像超解析度、視覺語言模型、退化表徵、擴散模型、Blind Image Super-Resolution、Vision Language Model、Degradation Representation、Diffusion Model

Leveraging Knowledge Distillation on Multi-Exit Vision Transformer for Small-Data Image Classification

電機資訊國際博士學位學程・共同指導：鄭文皇

視覺轉換器、自組學習、殘留標記嵌入、影像分類、知識提煉、Vision transformer、Self-ensemble learning、Residual token embedding、Image classification、Knowledge distillation

FreeLit: Paired-Free Indoor Relighting via Physics-Guided Diffusion

資訊科學與工程研究所

可控式重照明、室內場景重照明、無成對學習、擴散模型、物理引導式學習、Controllable Relighting、Indoor Scene Relighting、Unpaired Learning、Diffusion Models、Physics-Guided Learning

IDAG-Edit: Multi-Object Video Editing via Instance-Decoupled Attention and Guidance

數據科學與工程研究所

影片編輯、文字生成影片之擴散模型、多物件、細粒度控制、時間一致性、Video Editing、Text-to-Video Diffusion、Multi-Object、Fine-Grained Control、Temporal Consistency

LR-BiDiffusion: Bidirectional Latent Diffusion for LiDAR-Radar Translation

數據科學與工程研究所

跨模態生成、光達與雷達轉換、擴散模型、聯合分佈學習、自主感知、Cross-Modal Generation、LiDAR-Radar Translation、Diffusion Models、Joint Distribution Learning、Autonomous Perception

Animatable Head Avatar Reconstruction from Sparse Views via Triplane-conditioned Implicit Gaussian Field

數據科學與工程研究所

可控制3D頭部、3D高斯潑濺、稀疏視角重建、隱式高斯場、三平面表示、擴散式影像修復、Animatable 3D Avatar、3D Gaussian Splatting、Sparse-View Reconstruction、Implicit Gaussian Field、Triplane Representation、Diffusion-based Image Refinement

Text-Pilot: Intelligent Visual Text Planning and Manipulation via Multi-modal LLM as Agent

資訊學院碩士在職專班

文字轉圖片生成模型、視覺文字生成、視覺文字編輯、大型語言模型代理、ext To Image Generation Model、Visual Text Generation、Visual Text Editing、Large Language Model Agent

DNA: Dual-branch Network with Adaptation for Online Handwriting Generation

智能系統研究所

手寫筆跡生成、零樣本生成、手寫風格模仿、風格遷移、Online Handwriting Generation、Zero-Shot Generation、Handwriting Style Imitation、Style Transfer

UI-Styler: Ultrasound Image Style Transfer with Class-Aware Prompts for Cross-Device Diagnosis Using a Frozen Black-Box Inference Network

電機資訊國際碩士學位學程

風格轉換、提示微調、遷移學習、領域適應、Style Transfer、Prompt Tuning、Transfer Learning、Domain Adaptation

LAMP: Leveraging Autoregressive Modeling and Prompt Retrieval for Zero-Shot Time Series Forecasting

電機資訊國際碩士學位學程

時間序列預測、大型語言模型、基礎模型、提示微調、多模態、Time Series Forecasting、Large Language Models、Foundation Models、Prompt Tuning、Multi Modal

Blur2Sharp: Human Novel Pose and View Synthesis with Generative Prior Refinement

電機工程學系・共同指導：帥宏翰

新視角生成、新姿勢生成、神經渲染、擴散模型、單張影像人體生成、Novel Pose Synthesis、Novel View Synthesis、Neural Rendering、Diffusion Models、Single-Image Human Generation

Warm-up Strategies for Zero-Cost Differentiable NAS in Hardware Application Optimization

機器人碩士學位學程

神經網路架構搜索、物件偵測、電腦視覺、免訓練、硬體感知、Neural Architecture Search、Object Detection、Computer Vision、Training Free、Hardware-efficient

DynFaceRestore: Balancing Fidelity and Quality in Diffusion-Guided Blind Face Restoration with Dynamic Blur-Level Mapping and Guidance

智能系統研究所

擴散模型、盲臉恢復、分類器引導、Diffusion models、Blind Face Restoration、Classifier Guidance

Enhancing HVAC Control Stability and Generalization through Temporal Disentanglement with Multi-domain Dataset

數據科學與工程研究所

人工智慧、空調系統、預測控制、時間解耦、遷移學習、Artificial Intelligence、HVAC、Predictive Control、Temporal Disentanglement、Transfer Learning

Swiping between Hot and Cold Diffusion: Revisiting the Design of Diffusion Space with Blur-to-Noise Ratio in Image Generation

資訊科學與工程研究所

擴散機率模型、影像生成、資料流形、頻譜依賴性、分治策略、Diffusion probabilistic models、Image generation、Data manifold、Spectral dependency、Divide-and-conquer strategy.

RC-AutoCalib：An End-to-End Radar-Camera Automatic

智能系統研究所

校正、無標靶、雷達、相機、雙視圖表示、Calibration、Target-less、Radar、Camera、Dual-view Representation

DetailSemNet: Elevating Signature Verification with Captured Details and Semantics by Feature Disentanglement and Re-entanglement

人工智慧技術與應用碩士學位學程・共同指導：帥宏翰

離線簽名驗證、特徵解耦、局部匹配、Offline Signature Verification、Feature Disentanglement、Local Matching

MMM-TVR: Many-to-Many Matching for Text-Video Retrieval in a Novel Probabilistic Basic Concept Space

電機資訊國際碩士學位學程・共同指導：帥宏翰

視覺語言任務、文字影片檢索、表徵學習、半監督對比學習、軟偽標籤、Vision-Language Task、Text-Video Retrieval、Representation Learning、Semi-supervised Contrastive Learning、Soft Pseudo-labels

SOPP: Transfer Learning with Scene-Oriented Prompt Pool on 3D Object Detection

多媒體工程研究所

遷移學習、微調、提示微調、提示池、三維目標檢測、Transfer Learning、Fine Tuning、Prompt Tuning、Prompt Pool、3D Object Detection

Fast Dynamic Neural Radiance Fields

電機資訊國際碩士學位學程・共同指導：彭文志

深度學習、電腦視覺、新穎視圖合成、NeRF、Deep learning、Computer vision、Novel view syn- thesis

Improving Deep-Learning Electrocardiogram Classification Using Coloring Method

資訊科學與工程研究所・共同指導：曾建超

心房顫動、深度學習、人口統計資訊、心電圖著色、特徵著色、PTB-XL 心電圖數據資料集、Atrial fibrillation、Deep learning、Demographic information、ECG colorization、Feature coloring、PTB-XL

DA²: Degree-Accumulated Data Augmentation on Point Clouds with Curriculum Dynamic Threshold Selection

智能系統研究所

點雲數據增強、課程式動態閾值、點雲分類、師生模型、Degree-Accumulate Data Augmentation、Point Cloud Classification、Curriculum Dynamic Thresholding、Teacher-Student Model

Blind Super Resolution with Reference Images and Implicit Degradation Representation

資訊學院碩士在職專班

單一影像超解析度、盲超解析度、零樣本學習、參考圖像的影像超解析、非配對參考圖像的影像超解析、Image Super Resolution、Blind Super Resolution、Zero-shot Training、Reference-Based Super Resolution、Unpaired Reference-Based Super Resolution

Underwater Species-constrained Fish Detection using Multi-frame Image Information

電機資訊國際碩士學位學程・共同指導：杭學鳴

YOLOv5、深度學習、魚類檢測、輔助系統、運動歷史圖像、多幀圖像、deep learning、fish detection、auxiliary system、Motion History Image (MHI)、multi-frame image

Loss Reweight to Select Data by Data Importance For 3D Object Detection

網路工程研究所

光達、三維物件偵測、樣本選擇、損失函數調整、LiDAR、3D Object detection、sample selection、loss reweight

IZero-DNAS - Initialization independent Zero cost Differentiable Neural Architecture Search for object detection

多媒體工程研究所

神經網路架構搜索、零消耗指標、物件爭測、Neural Architecture Search、Zero-Cost Proxy、Object Detection

Adaptive Model-Based Transfer Learning for Dynamic HVAC Control

人工智慧技術與應用碩士學位學程

大數據分析、暖通空調系統、智慧建築、遷移學習、Big data analysis、HVAC、smart building、transfer learning (TL)

Image Blind Super-Resolution Based on Implicit Prediction and Decomposition of Multiple Degradation Factors

多媒體工程研究所・共同指導：莊仁輝、邱維辰

盲超解析度、對比學習、無監督學習、解開表示學習、雙分支網絡、blindSR、contrastive learning、unsupervised learning、distangled representation learning、dual branch learning

Few-Shot Generalizable Time-Dependent Neural Radiance Fields for Synthesizing Novel Views at Different Times

資訊科學與工程研究所

新視角生成、隱式神經表示、神經渲染、少量樣本、泛化、時間依賴生成、novel view synthesis、implicit neural representation、neural rendering、few-shot、generalization、time-dependent synthesis

Fusion UandV: Efficient MRI Sampling via the Fusion of U-Net and Conditional Variational Autoencoder

電機資訊國際碩士學位學程

加速磁共振成像 (MRI)、逆反問題、深度學習、電腦計算機視覺、Accelerated Magnetic Resonance Imaging (MRI)、Inverse problem、Deep Learning、Computer Vision

Zero-Reference End-to-End Network for Low-light Image Enhancement with Noise Reduction

資訊科學與工程研究所

零參考、無監督學習、影像處理、低光影像增強、影像降噪、Zero-reference、Unsupervised learning、Image processing、Low-light image enhancement、Image denoising

Implicit Image Function Modeling for Arbitrary-Resolution and Arbitrary-Scale Face Super-Resolution

資訊科學與工程研究所

人臉超解析度、任意倍率超解析度、隱式神經表示、任意解析度與任意倍率、全局引導、face super-resolution、arbitrary-scale super-resolution、implicit neural representation、arbitrary-resolution and arbitrary-scale、global guidance

Towards Understanding Cross Resolution Feature Matching for Surveillance Face Recognition

網路工程研究所

跨解析度人臉辨識、監視器人臉辨識、超解析度、Cross-resolution face recognition、Surveillance face recognition、Super-resolution

ExReg: Wide-range Photo Exposure Correction via Regressor

資訊科學與工程研究所

影像強化、影像處理、影像曝光校正、多維度迴歸、Image Enhancement、Image Processing、Image Exposure Correction、Multi-dimension Regression

LiDAR and Image-based Perception System for Autonomous Driving

多媒體工程研究所

自駕車、深度學習、物件偵測、感知系統、self-driving、deep learning、object detection、perception system

MENTOR: Multilingual tExt detectioN TOward leaRning by analogy

網路工程研究所

多國語言、文字偵測、動態內核、multi-lingual、text detection、dynamic kernel

Aerogel : Incremental map modeling for Lightweight SLAM via Deep Reinforcement Learning

資訊學院資訊學程

SLAM、關鍵幀、深度強化學習、keyframe、deep reinforcement learning

Efficient constraint-aware neural architecture search for object detection

電機資訊國際學程

神經網路架構搜索、物體探測、零耗網路重要性指標、深度學習、計算機視覺、Neural Architecture Search (NAS)、Object Detection、Zero-Cost Proxy、Deep Learning、Computer Vision

Anomaly Detection with Neural Basis Analysis and InfoNCE on Residential Electricity Load Data

多媒體工程研究所

深度學習、時序資料、異常偵測、基底分析、資訊噪聲對比估計、deep learning、time series data、anomaly detection、Nbeats、InfoNCE

SSIOE: Self-Supervised Indoor Occupancy Estimation For Intelligent Building Management

智慧系統與應用研究所・共同指導：林柏宏

占用估計、深度學習方法、環境監測與管理、感測器融合、基於感測器的控制、自監督式學習、元學習、Occupancy Estimation、Deep Learning Methods、Environment Monitoring and Management、Sensor Fusion、Sensor-based Control、Self-Supervised Learning、Meta-Learning

INDIRECT: Invertible and Discrete Noisy Image Rescaling with Enhancement from Case-dependent Textures

資訊科學與工程研究所

雜訊影像重縮放、影像重縮放、離散可逆神經網路、案例相關的紋理資訊網路、Noisy image rescaling、Image rescaling、Discrete invertible neural network、Case-dependent texture network

Find the Way Back: Invertible Kernel Estimator for Zero-shot Blind Image Super Resolution

資訊科學與工程研究所・共同指導：邱維辰

單一影像超解析度、零樣本學習、流生成模型、自我監督學習、Single Image Super Resolution、Zero-shot Learning、Flow-based Generative Model、Self-supervised Learning

Density-imbalance-eased LiDAR Point Cloud Upsampling via Feature Consistency Learning

多媒體工程研究所

點雲、三維表現特徵、流型升取樣、三維重建、自駕車、Point Cloud、3D Representation、Manifold-based Upsampling、3D Reconstruction、Autonomous Vehicle

SVDnet: Singular Value Control and Distance Alignment Network for 3D Object Detection

多媒體工程研究所

點雲、物件偵測、自駕車、Point Cloud、3D object detection、Autonomous Vehicle

Decentralized Image Deployment in Datacenter

資訊科學與工程研究所・共同指導：黃世昆

點對點、服務開通、去中心化、Peer-to-Peer、BitTorrent、System Provisioning