Deep Domain Adaptation: Cross-Modality Visual Perception, Synthesis, and Beyond
深度適應學習:跨視覺資料模組之感知、合成與前瞻應用

National Taiwan University Prof. Yu-Chiang Frank Wang

計畫主持人:國立臺灣大學電機工程學系 王鈺強教授

以轉移學習與領域適應為主軸,配合目前逐漸受矚目應用的智慧機器人與各式視覺元件,發展跨視覺資料領域之辨識、分析與合成之技術,達到提升使用者生活品質,並促進技術落地,更達到促進國內相關AI產業人才之目的。

Focusing on transfer learning and domain adaptation, the project aims to develop cross-domain technologies of identification, analysis and synthesis with various visual components that are gradually being used. Furthermore, the project also aims to improve the quality of users’ lives and cultivate talents in the AI industry.