Chop & Learn: Recognizing and Generating Object-State Compositions
Chop & Learn: Recognizing and Generating Object-State Compositions
复制标题
DOI：
10.1109/iccv51070.2023.01852
复制
发表时间：
2023-09
期刊：
影响因子：
--
通讯作者：
Nirat Saini;Hanyu Wang;Archana Swaminathan;Vinoj Jayasundara;Bo He;Kamal Gupta;Abhinav Shrivastava
中科院分区：
文献类型：
--
作者：
Nirat Saini;Hanyu Wang;Archana Swaminathan;Vinoj Jayasundara;Bo He;Kamal Gupta;Abhinav Shrivastava
识别和生成对象状态组合一直是一项具有挑战性的任务，特别是在推广到看不见的组合时。在本文中，我们研究的任务，切割对象在不同的风格和由此产生的对象状态的变化。我们提出了一个新的基准套件Chop & Learn，以适应学习对象和使用多个视点的不同切割风格的需求。我们还提出了一个新的任务的合成图像生成，它可以通过生成新的对象状态图像，将学习到的剪切样式转移到不同的对象。此外，我们还将视频用于组合动作识别，并展示了该数据集在多个视频任务中的宝贵用途。项目网址：https://chopnlearn.github.io。
Recognizing and generating object-state compositions has been a challenging task, especially when generalizing to unseen compositions. In this paper, we study the task of cutting objects in different styles and the resulting object state changes. We propose a new benchmark suite Chop & Learn, to accommodate the needs of learning objects and different cut styles using multiple viewpoints. We also propose a new task of Compositional Image Generation, which can transfer learned cut styles to different objects, by generating novel object-state images. Moreover, we also use the videos for Compositional Action Recognition, and show valuable uses of this dataset for multiple video tasks. Project website: https://chopnlearn.github.io.