ReMoMask-2: Latent Retrieval-Augmented Masked Motion Generation

  • 类型:arxiv
  • 标识:2609.08365
  • 链接:https://arxiv.org/abs/2609.08365
  • 主分类:rag
  • 形态:method
  • 被引:0
  • 被引来源:Semantic Scholar + OpenAlex
  • S2被引:0
  • OpenAlex被引:0
  • 影响力被引:0
  • TLDR:ReMoMask-2, a structure-aware RAG framework coupling Hierarchical Bidirectional Momentum contrastive learning to align global and part-level features with text; Semantic Spatial-Temporal Attention (SSTA) for topology-aware fusion; and Topology Structured Masking (TSM) to force robust part-level grounding via adaptive masking are introduced.
  • OpenAlex ID:W7212035826
  • OpenAlex DOI:10.48550/arxiv.2609.08365
  • DOI:10.48550/arxiv.2609.08365
  • DOI来源:OpenAlex
  • 开放获取:green
  • 开放获取链接:https://arxiv.org/pdf/2609.08365
  • OpenAlex更新:2026-09-16
  • 待LLM分类:否
  • 标题中文:ReMoMask-2:潜在检索增强的掩码运动生成
  • TLDR中文:提出结构感知的 RAG 框架 ReMoMask-2:耦合 Hierarchical Bidirectional Momentum 对比学习以对齐全局与部件级特征与文本;采用 Semantic Spatial-Temporal Attention (SSTA) 实现拓扑感知的融合;通过 Topology Structured Masking (TSM) 借助自适应掩码强化鲁棒的部件级 grounding。
  • 来源文件:
  • /inbox/tom/_candidates/2026-09-15-rag-retrieval-reranking-candidates.json
  • [OpenAlex backfill]
  • [S2 enrich]