Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment 文章

ArXiv CS.CV2026-07-07PAPERen作者: Zixiang Zhou, Zhentao Yu, Yifeng Ma, Hongmei Wang, Wenqing Yu, Cong Wang, Zilin Yang, Rui Chen, Jiarong Ou, Yezhou Liu, Yuan Zhou, Qinglin Lu
大语言模型
查看原文 →
关系图谱

Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment · 相关技术

相关技术

CAPACTAbstractACEABLEAHATAI-based Visually Impaired AssistanceANNANSIAR video generationATEAdaptive Batch ScalingAdvanced Persistent ThreatsArabic Speech Emotion RecognitionAuRAAudio Mamba EncoderAugmented Latent IntrinsicsBRIDGEBotCOMCOVCOVERCTICTUCalibrated Entropy ScoreClinical Trust SignalsDELDESDIFDIFFDIRDIRECTDIffusion Reconstruction ErrorDeep active learningDenseDiTDiffusion Transformer