English(EN) Improving Long-Tail Detection via Layout-Instruction Conditioned Diffusion Synthesis

Tail-SD 使用扩散合成改进稀有物体检测

作者 PulseAugur 编辑部 · [1 个来源] · 2026-11-05 07:00

研究人员开发了 Tail-SD，一个使用扩散模型生成合成数据以改进物体检测（特别是稀有类别）的新颖框架。该方法允许使用文本、布局或两者的组合进行可控合成，解决了先前仅文本或仅布局方法的局限性。在 COCO 和 LVIS 数据集上的实验显示，稀有物体的检测得到了显著改进，稀有类别的 AP 显著增加。 AI

影响这项研究为生成合成数据以改进物体检测（特别是稀有类别）提供了一种新方法，这可能有利于依赖于准确物体识别的 AI 系统。

排序理由该集群描述了一篇提出新物体检测方法的论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 Hugging Face Daily Papers 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群描述了一篇提出新物体检测方法的论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, model release; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: 5 days old; Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

Hugging Face Daily Papers TIER_1 English(EN) · 2026-11-05 07:00

通过布局指令条件化扩散合成改进长尾检测

Synthetic data from diffusion models can improve object detection, but long-tail performance is still limited by scarce rare classes and inflexible conditioning (e.g., text-only or layout-only inputs). In this paper, we propose Tail-SD, a layout-instruction conditioned stable dif…

本文如何被排名

报道来源 [1]

通过布局指令条件化扩散合成改进长尾检测

相关实体

相关话题