極簡筆記 Fully Convolutional Adaptation Networks for Semantic Segmentation

極簡筆記 Fully Convolutional Adaptation Networks for Semantic Segmentation 文章核心貢獻,提出FCAN,探究利用GTA5遊戲數據集來訓練語義分割網絡,並將此網絡遷移到真實路況場景下進行測試。 FCAN分爲兩部分,Appearance Adaptation Networks(ANN)和Representation Adaptation
相關文章
相關標籤/搜索