Clipart1k

Introduced by Inoue et al. in Cross-Domain Weakly-Supervised Object Detection through Progressive Domain Adaptation

In Clipart1k, the target domain classes to be detected are the same as those in the source domain. All the images for a clipart domain were collected from one dataset (i.e., CMPlaces) and two image search engines (i.e., Openclipart2 and Pixabay3). Search queries used are 205 scene classes (e.g., pasture) used in CMPlaces to collect various objects and scenes with complex backgrounds.

Homepage

Benchmarks

Add a new result Link an existing benchmark

Task	Dataset Variant	Best Model
Weakly Supervised Object Detection	Clipart1k	H2FA R-CNN
Body Detection	Clipart1k	DASS-Detector
Unsupervised Object Detection	Clipart1k	CutLER
Object Detection	Clipart1k	CDDMSL