Tell us about your project!
DATASET
The dataset solves computer vision tasks for training neural networks in photo and video editing Apps
Dancing People Segmentation
Computer Vision
Segmentation
Splitting an image into semantically independent segments to define the boundaries of objects
The ability to recognize and analyze images and video
Using machine learning in augmented reality applications to create filters and effects
Data Annotation
Labeling objects in photos to train systems to recognise and interpret them
AR
2 500
images
2 weeks
project duration
8.75
kaggle usability
Case description:
Original image
The video sequence with dancing people was divided by frames into 2615 images
For each image created a mask in PhotoShop with high-quality markup in 97%
Formed collages: original photo - only background - only person
Case description:
Mask: background without object
The video sequence with dancing people was divided by frames into 2615 images
For each image created a mask in PhotoShop with high-quality markup in 97%
Formed collages: original photo - only background - only person
Case description
Mask: background without object
The video sequence with dancing people was divided by frames into 2615 images
For each image created a mask in PhotoShop with high-quality markup in 97%
Formed collages: original photo - only background - only person
Tell us about your project!
Data collection and annotation services:
Landmark: marking the keypoint of the positions and location of the object in space
Semantic segmentation of objects in the image by a set of attributes
Data collection from various open sources using parsing or private photos of people using crowdsourcing.
Image classification and tagging
Tell us about your project!
See other datasets
