Generated Contents Enrichment

  • 类型:arxiv
  • 标识:2405.03650
  • 链接:https://arxiv.org/abs/2405.03650
  • 主分类:multimodal
  • 形态:method
  • 被引:1
  • 被引来源:Semantic Scholar
  • S2被引:1
  • OpenAlex被引:0
  • 影响力被引:0
  • TLDR:A jointly trained adversarial framework is proposed that enriches scene graphs by modeling object semantics and inter-object relations and is evaluated with proxy scene graph enrichment metrics, image-quality comparisons, qualitative examples, and user studies on the Visual Genome dataset.
  • OpenAlex ID:W4396786651
  • OpenAlex DOI:10.48550/arxiv.2405.03650
  • DOI:10.48550/arxiv.2405.03650
  • DOI来源:OpenAlex
  • 开放获取:green
  • 开放获取链接:https://arxiv.org/pdf/2405.03650
  • OpenAlex更新:2026-07-19
  • 待LLM分类:否
  • 成熟度:research
  • 场景:scene graph enrichment、image understanding、adversarial training
  • 标题中文:生成内容增强
  • TLDR中文:本文提出一个联合训练的对抗框架,通过建模对象语义和对象间关系来增强场景图,并在 Visual Genome 数据集上以代理场景图增强指标、图像质量比较、定性示例与用户研究进行评估。
  • 来源文件
  • /inbox/tom/_candidates/2026-07-07-rag-retrieval-reranking-candidates.json
  • [S2 enrich]
  • [OpenAlex backfill]