Create Trainset for SmartTool
Prepare training data for SmartTool
Transform data and annotations, perform augmentations, filtering and querying…
Prepare training data for SmartTool
Converts shapes of classes (e.g. polygon to bitmap) and all corresponding objects
Creates images project from video project
Assigns tags (train/val) to images. Training apps will use these tags to split data.
Generate synthetic data: flying foregrounds on top of backgrounds
Merge multiple classes with same shape to a single one
Visualize and build augmentation pipeline with ImgAug
for both images and their annotations
Configure, preview and split images and annotations with sliding window
Read every n-th frame and save to images project
Visual diff and merge tool helps compare project tags and classes
Convert classes to bitmap and rasterize objects without intersections
Visual diff and merge tool helps compare images in two projects
Merge selected datasets with images or videos into a single one
Tags and object classes can be customized
Generate synthetic data for classification of retail products on grocery shelves
Synthesize videos on annotated data
Split "AnyShape" classes to classes with strictly defined shapes (polygon, bitmap, ...)
Creates new project with cropped objects
Creates video project from images project
Merge images and labels that were split by sliding window before
Create foreground mask from alpha channel of image
Converts shapes of classes on videos (e.g. polygon to bitmap) and all corresponding objects
Creates video from images in dataset with selected frame rate and configurable label opacity
Saves tag to images mapping to a json file
Rename "Key:Value" tags to key_value (fruit: lemon -> fruit_lemon)
Creates sequence of connected point clouds with tracklets
Convert each class name to tag associated with objects, and merge existing classes into single one
Create new object classes from tags associated with objects
Change video framerate with preserving duration (recodes video)