EditCaption: Human-Refined SFT and HAE-DPO for Image Editing Instruction Synthesis 文章

ArXiv CS.CV2026-05-26NEWSen作者: Xiangyuan Wang, Honghao Cai, Yunhao Bai, Chao Hui, Tianze Zhou, Haohua Chen, Hao Shi, Yuling Wu, Yao Hu, Xu Tang, Yibo Chen, Wei Zhu
大语言模型
查看原文 →
关系图谱

EditCaption: Human-Refined SFT and HAE-DPO for Image Editing Instruction Synthesis · 相关技术

相关技术

ORMODEPLAlanguage model远程代码执行（RCE）研究法辛烷值后训练evalUCTStraight-Through EstimatorSSESFTReferring expression comprehension (REC)Parts-of-Speech (POS) tagsVision-Language ModelsVLMHISForEffort Metric AttentionEFIDiTDPOCISARGANNTable