Language
The catalogue lists 10 Nomatsiguenga datasets. Most listings are image, video and medical data, typically used for document understanding & OCR.
Image
A dataset of both selfie pictures and pictures of the ID of an official identity document.
Video
[a dataset of 1 or 2 180 selfie videos per person.
Video
This dataset contains more than 3000 videos of partially or fully naked people, with detailed annotations of body parts in various moments of each video.
Image
A collection of up to 8 selfies per person and up to 2 pictures of an ID (ID picture only), distributed over a variety of ethnicities, skin tones, ages and genders.
Image
A collection of images of drugs and drug use, ideal for content moderation purposes.
Image · Medical
A collection of images of surgery, injuries, wounds, diseases and similar content, ideal for content moderation purposes.
Image
A collection of images of gory scenes, human bodies and similar content, ideal for content moderation purposes.
Image
A collection of images of sexual nature, ranging from occluded nudity, nudism and racy images, to pornographic images; ideal for content moderation purposes.
Image
A collection of images containing hate symbols, lethal weapons, and terrorism; ideal for content moderation purposes.