NoOVD：面向开放词表目标检测的新类别发现与嵌入
NoOVD: Novel Category Discovery and Embedding for Open-Vocabulary Object Detection
利用冻结视觉语言模型的知识进行自蒸馏，引导RPN和RoI头发现新类别目标，避免被误判为背景；并通过R-RPN在推理时调整候选框置信度，提升新类别召回率。
原文依据 we propose a novel training framework--NoOVD--which innovatively integrates a self-distillation mechanism grounded in the knowledge of frozen vision-language models (VLMs).
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论尽管开放词表目标检测（OVD）已取得显著进展，但训练阶段与测试阶段之间仍存在明显差距。训练时，RPN和RoI头常将未标注的新类别目标误判为背景，导致部分候选框被RPN过早过滤，另一些则被RoI头进一步误分类。测试时，这些候选框再次获得低分并在后处理中被移除，造成召回率显著下降，最终削弱新类别检测性能。为解决这些问题，我们提出一种新的训练框架——NoOVD，创新性地集成了基于冻结视觉语言模型（VLM）知识的自蒸馏机制。具体而言，我们设计了K-FPN，利用VLM的预训练知识引导模型发现新类别目标并促进知识蒸馏——无需额外数据——从而防止新类别目标被强制对齐为背景。此外，我们引入R-RPN，在推理时调整候选框的置信度分数，以提高新类别目标的召回率。在OV-LVIS、OV-COCO和Objects365上的跨数据集评估表明，我们的方法在多个指标上持续取得更优性能。
英文原文摘要 1345 字符
Despite the remarkable progress in open-vocabulary object detection (OVD), a significant gap remains between the training and testing phases. During training, the RPN and RoI heads often misclassify unlabeled novel-category objects as background, causing some proposals to be prematurely filtered out by the RPN while others are further misclassified by the RoI head. During testing, these proposals again receive low scores and are removed in post-processing, leading to a significant drop in recall and ultimately weakening novel-category detection performance.To address these issues, we propose a novel training framework--NoOVD--which innovatively integrates a self-distillation mechanism grounded in the knowledge of frozen vision-language models (VLMs).Specifically, we design K-FPN, which leverages the pretrained knowledge of VLMs to guide the model in discovering novel-category objects and facilitates knowledge distillation--without requiring additional data--thus preventing forced alignment of novel objects with background.Additionally, we introduce R-RPN, which adjusts the confidence scores of proposals during inference to improve the recall of novel-category objects.Cross-dataset evaluations on OV-LVIS, OV-COCO, and Objects365 demonstrate that our approach consistently achieves superior performance across multiple metrics.