Introduction to the PartImageNet Semantic Part Segmentation dataset

PartImageNet is a paper published in ECCV2022, which proposes an object part-level annotation dataset with the largest number of current categories and images.

PartImageNet: A Large, High-Quality Dataset of Parts

paper: https://arxiv.org/abs/2112.00933

code: https://github.com/TACJu/PartImageNet

The data set download link has been released on GitHub and can be downloaded.

---

The PartImageNet dataset contains 158 categories and a total of 24,095 images. Each image contains part-level annotations for a single object. An example of the annotation is shown in the figure below.

The specific category information of the dataset is shown in the table below. The 158 subcategories belong to 11 categories. The number of subcategories in each category is given in parentheses.The labeled part classes are the same.Taking the Fish category as an example, it contains 10 sub-categories, and the part categories of objects in each category are Head, Body, Fin, and Tail.

Out of the 158 classes, 118 are non-rigid body classes (eg dogs) and 40 are rigid body classes (eg cars).In addition, the dataset also provides more fine-grained classification information, as shown in the figure below.Example: Quadruped → Dog → Gordon setter.

---

For the Semantic Part Segmentation task, the data set is divided according to 85%, 5%, and 10%. The specific information is as follows:

By the way, let's take a look at the performance indicators of existing methods on this task, as follows:

---

For the Few-shot Learning task, the data set is divided according to the number of categories 109, 19, and 30. The specific information is as follows: