语言条件检测Transformer
Language-conditioned Detection Transformer
AI总结：
本文提出DECOLA框架，通过先训练语言条件检测器生成高质量伪标签，再训练无条件开放词汇检测器，在多个零样本检测基准上大幅超越现有方法。
摘要：
我们提出了一种新的开放词汇检测框架。该框架在可用时同时使用图像级标签和详细的检测标注。我们的框架分三步进行。我们首先在完全监督的检测数据上训练一个语言条件目标检测器。该检测器在训练过程中能够看到真实类别的存在或缺失，并以存在的类别集合为条件进行预测。我们使用该检测器为带有图像级标签的图像生成伪标签。凭借其条件机制，我们的检测器提供的伪标签比先前方法准确得多。最后，我们在伪标注图像上训练一个无条件开放词汇检测器。所得检测器名为DECOLA，在开放词汇LVIS基准以及LVIS、COCO、Object365和OpenImages上的直接零样本迁移基准中表现出强大的零样本性能。DECOLA在零样本LVIS基准上比先前方法高出17.1 AP-rare和9.4 mAP。DECOLA仅通过在开源数据和学术规模计算上训练，就在各种模型规模、架构和数据集上取得了最先进的结果。代码可在https://github.com/janghyuncho/DECOLA获取。
英文摘要：
We present a new open-vocabulary detection framework. Our framework uses both image-level labels and detailed detection annotations when available. Our framework proceeds in three steps. We first train a language-conditioned object detector on fully-supervised detection data. This detector gets to see the presence or absence of ground truth classes during training, and conditions prediction on the set of present classes. We use this detector to pseudo-label images with image-level labels. Our detector provides much more accurate pseudo-labels than prior approaches with its conditioning mechanism. Finally, we train an unconditioned open-vocabulary detector on the pseudo-annotated images. The resulting detector, named DECOLA, shows strong zero-shot performance in open-vocabulary LVIS benchmark as well as direct zero-shot transfer benchmarks on LVIS, COCO, Object365, and OpenImages. DECOLA outperforms the prior arts by 17.1 AP-rare and 9.4 mAP on zero-shot LVIS benchmark. DECOLA achieves state-of-the-art results in various model sizes, architectures, and datasets by only training on open-sourced data and academic-scale computing. Code is available at https://github.com/janghyuncho/DECOLA.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
