謝君偉

Semantic Discrimination Framework with Negative Text Examples for Zero-Shot Object Counting

智慧系統與應用研究所

零樣本物件計數、影像相依的負向提示詞、偽陽性抑制、zero-shot object counting、image-dependent negative prompt、false-positive suppression

Clone-based Neural Architecture Search with Adaptive Feature Interaction for 3D Medical Image Segmentation

智慧計算與科技研究所

三維醫學影像分割、神經架構搜尋、自適應特徵互動、Medical Image Segmentation、Neural Architecture Search、Adaptive Feature Interaction

GPS-denied Drone Navigation via Cross-Domain Keypoint Matching

智慧系統與應用研究所

特徵點匹配、跨域影像、無人機導航、電腦視覺、深度學習、Keypoint Matching、Cross-domain Imagery、Unmanned Aerial Vehicle Navigator、Computer Vision、Deep Learning

Enhancing DETR with YOLO-Inspired Object Detectors

智慧系統與應用研究所

物件偵測、YOLO、電腦視覺、深度學習、Object detection、Computer Vision、Deep Learning

SALA-Net: Scale-Adaptive and Layer-Aware Network for 3D Medical Image Segmentation

智慧計算與科技研究所

醫學影像分割、狀態序列模型、深度學習、Medical Image Segmentation、State-space Models、Deep Learning

TISA-GS: Text-Guided Object Insertion and Shadow-Aware Relighting in Gaussian Splatting Scenes

智慧計算與科技研究所

3D高斯模型、3D編輯、電腦視覺、深度學習、3D Gaussian、3D Edit、Computer Vision、Deep Learning

Building Detection Hybrid Architecture Based on Deep Learning

智慧與綠能產學研究所

建築物偵測、影像切割、物件偵測、分類器、深度學習、Building detection、Segmentation、Object detection、Classifier、Deep Learning

Enhancing Robust Federated Learning for Long-Tailed Distributions with Noisy Clients

智慧系統與應用研究所

穩定聯邦式學習、聯邦式長尾學習、聯邦式噪聲標籤學習、電腦視覺、深度學習、Robust Federated Learning、Federated Long-Tail Learning、Federated Noisy Label Learning、Computer Vision、Deep Learning

CSPA: Unsupervised Point Cloud Semantic Segmentation via Multi-Scale Superpoint Attention and Refinement

智慧計算與科技研究所

點雲、語義分割、無監督學習、多尺度注意力機制、超點、Point cloud、Semantic Segmentation、Unsupervised learning、Cross-scale attention、Superpoint

MUST-MAS: MUlti-Scale aTtention with Macro Architecture Search Model for Medical Image Segmentation

智慧計算與科技研究所

電腦視覺、深度學習、醫學影像、Computer Vision、Deep Learning、Medical Imaging

Channel-aware Multi-scale Efficient Partial Recursive Attention for Vision Transformer

智慧計算與科技研究所

輕量化模型、電腦視覺、深度學習、Efficient Vision Transformer、Computer Vision、Deep Learning

MAESTRO: Multi-Anchor and Dual-Constraint Enhanced Sequential Training for Robust Optimization in Image Editing

智慧計算與科技研究所

穩定擴散模型、生成式 AI、圖像編輯、電腦視覺、深度學習、Stable Diffusion、Generative Artificial Intelligence、Image Eidt、Computer Vision、Deep Learning

Addressing Long-Tail Noisy Label Learning Problems: a Two-Stage Solution with Label Refurbishment Considering Label Rarity

智慧系統與應用研究所

長尾學習、噪音標籤、分類、電腦視覺、深度學習、Long-tailed Learning、Noisy Label、Classification、Computer Vision、Deep Learning

Enhancing Visual Detection with Mixed Stage Partial Network, Parallel FPN, and Background Learning

資訊科學與工程研究所・共同指導：陳永昇

小物件偵測、特徵金字塔、資料擴增、無人機、智慧運輸系統、small object detection、feature pyramid network、data augmentation、unmanned aerial vehicle、intelligent transportation system

Patch-based Cross-scale Attention Network for Oil Spill and Crack Detection

智慧計算與科技研究所

電腦視覺、深度學習、油污偵測、路損偵測、跨尺度注意力模塊、Computer Vision、Deep Learning、Oil Spill Detection、Crack Detection、Cross-scale Attention Module

Feature Shuffle Network with Central line and Transformer for Scene Text Detection

智慧系統與應用研究所

場景文字偵測、小場景文本、主幹、電腦視覺、深度學習、Scene Text Detection、small scene text、backbone、Computer Vision、Deep Learning

Cooperative Reinforcement Learning for Traffic Signal Control and Carbon Emission Reduction

智慧與綠能產學研究所

深度學習、紅綠燈控制、全域合作、交通控制、碳排放、Deep Learning、Traffic Light Control、Global Cooperation、Traffic Control、Carbon Emission.

Fine-tuning with Intra-Block Fusion and Cross-scale Atten-tion can make a difference in Few-Shot Learning

智慧系統與應用研究所

少樣本學習、電腦視覺、深度學習、注意力模塊、影像分類、Few-Shot Learning、Computer Vision、Deep Learning、Attention module、Image Classification

Latency-Aware Neural Architecture Search With Estimated Latency Based on Complexity and MAC

智慧計算與科技研究所

神經網路架構搜索、延遲感知神經網路架構搜索、神經網路複雜性、SmoothDARTS、BRP-NAS、Neural Architecture Search、latency-aware NAS、neural network complexity

SMILEtrack: SiMIlarity LEarning for Multiple Object Tracking

智慧與綠能產學研究所

電腦視覺、影像辨識、深度學習、多物件追蹤、Computer Vision、Multiple Object Tracking、Deep Learning、Object detection

NAS-based Messaged U-Nets with Reversible Pooling for Image Segmentation

智慧與綠能產學研究所

深度學習、醫學影像、神經網絡架構搜尋、語義切割、Deep Learning、Medical Image、Neural Architecture Search、Semantic Segmentation

SSS-Net: Spatial and Scaling Aware Siamese Network for Change Detection

智慧計算與科技研究所

電腦視覺、深度學習、影像辨識、變化偵測、注意力模塊、遙感探測、Computer Vision、Deep Learning、Change Detection、Attention module、Remote Sensing、Image Classification

Learnable Discrete Wavelet Pooling (LDW-Pooling) for Convolutional Networks

智慧系統與應用研究所

電腦視覺、深度學習、影像辨識、池化層、小波轉換、Computer Vision、Deep Learning、Image Classification、Pooling、Wavelet

Global Cooperation DDPG on Traffic Signal Control

智慧計算與科技研究所

機器學習、深度學習、強化學習、交通控制、Deep Learning、Reinforcement Learning、Traffic Signal Control

Chest X-ray Segmentation from Small Dataset by Data Augmentation Using GAN

智慧計算與科技研究所・共同指導：郭柏村

電腦視覺、深度學習、影像切割、資料擴增、生成對抗網路、Computer Vision、Deep Learning、Semantic Segmentation、Data Augmentation、Generative Adversarial Network

Novel Network with Interpolation Layer for scene text detection and recognition

智慧計算與科技研究所

場景文本檢測、場景文本辨識、Scene text detection、Scene Text Recognition

Real-time Crowd Counting Architecture

智慧計算與科技研究所

人群計數、熱點圖、輕量化、小物件、深度學習、Crowd Counting、Heatmap、Lightweight、Small-object、Deep Learning

Bi-Fusion Spatial Attention and Noise Data Augmentation for Semantic Segmentation

智慧計算與科技研究所

電腦視覺、語義分割、注意力模塊、融合尺度空間注意力、資料增強、Computer Vision、Semantic Segmentation、Attention Module、Bi-fusion Spatial Attention、Data Augmentation

Mean Shift Based Differential Architecture Search

智慧計算與科技研究所

電腦視覺、深度學習、影像辨識、神經架構搜索、可微分式架構搜索、Computer Vision、Deep Learning、Image Classification、NAS、DARTS