# 報告時程表

| 組別 | 時間 | 題目 |
| :--- | :--- | :--- |
| 1 | 09:00-09:15 | AI Adventurer: An Interactive Adventure Game Combining LLM-Based Storytelling and Human Action Recognition |
| 2 | 09:15-09:30 | AI 模型倫理與偏見實驗 |
| 3 | 09:30-09:45 | Circuit Element Identification Traditional SVM Approach versus YOLO Object Detection |
| 4 | 09:45-10:00 | A Multi-Agent System for Autonomous Narrative World-Building using LangGraph |
| | **10:00-10:10** | **休息時間** |
| 5 | 10:10-10:25 | Real-Time Visual Sketch Reproduction and AI Doodle Recognition System |
| 6 | 10:25-10:40 | Frequency-Aware Deepfake Detection |
| 7 | 10:40-10:55 | From Seeing to Doing: A Vision-Language Grounded Robotics System |
| 8 | 10:55-11:10 | 動物用藥小貼士 |
| 9 | 11:10-11:25 | CT-Guided PET Tumor Segmentation: Dual-Modal Fusion for False-Positive Suppression based on SAM-Med3D |
| 10 | 11:25-11:40 | Low-Light Image Enhancement Network Based on HVI Color Space |
| | **11:40-13:10** | **中午休息** |
| 11 | 13:10-13:25 | Detecting Poisoned Samples with Deep Abstaining Classifier |
| 12 | 13:25-13:40 | 智慧健康豬隻行為辨識系統 |
| 13 | 13:40-13:55 | AI生成圖中文字的完整化與再渲染工具 |
| 14 | 13:55-14:10 | 死而復聲-人工智慧語音聊天 |
| | **14:10-14:20** | **休息時間** |
| 15 | 14:20-14:35 | 基於VLM之車輛事故危險預測與場景語意敘述系統 |
| 16 | 14:35-14:50 | Efficient Edge Multimodal Biometrics via Hyperdimensional Computing |
| 17 | 14:50-15:05 | Object Detection Using Turtlebot3 in a Two-Line Track |
| 18 | 15:05-15:20 | Hybrid Memory Architecture for Scalable ViT Token Reduction on FPGAs |
| 19 | 15:20-15:35 | 觀察LLM在親自操作股票時的盈利狀況 |