论文标题
贴片对比度学习的开放词汇语义细分
Open Vocabulary Semantic Segmentation with Patch Aligned Contrastive Learning
论文作者
论文摘要
我们引入了贴片对比度学习（PACL），这是剪辑对比损失的修改兼容函数，打算训练视觉编码器的贴片令牌与文本编码器的CLS令牌之间的对齐。通过这样的对齐，模型可以识别与给定文本输入相对应的图像的区域，因此无缝地转移到开放词汇语义分割的任务中，而无需在训练过程中进行任何分割注释。使用PACL的预训练的剪辑编码器，我们能够在4种不同的分段基准上为开放词汇零摄像分段的任务设置最新的任务：Pascal VOC，Pascal Context，可可件，可可件和ADE20K。此外，我们表明PACL也适用于图像级预测，与夹子主链一起使用时，与夹子相比，在12个图像分类数据集的套件中，与夹子相比，零摄像机分类精度具有一般改善。
We introduce Patch Aligned Contrastive Learning (PACL), a modified compatibility function for CLIP's contrastive loss, intending to train an alignment between the patch tokens of the vision encoder and the CLS token of the text encoder. With such an alignment, a model can identify regions of an image corresponding to a given text input, and therefore transfer seamlessly to the task of open vocabulary semantic segmentation without requiring any segmentation annotations during training. Using pre-trained CLIP encoders with PACL, we are able to set the state-of-the-art on the task of open vocabulary zero-shot segmentation on 4 different segmentation benchmarks: Pascal VOC, Pascal Context, COCO Stuff and ADE20K. Furthermore, we show that PACL is also applicable to image-level predictions and when used with a CLIP backbone, provides a general improvement in zero-shot classification accuracy compared to CLIP, across a suite of 12 image classification datasets.
