Background
What is a data labeling tool?
Labeled datasets are essential for supervised machine learning. A data labeling tool is a system where annotators can create labeled datasets. It typically contains a user interface where the annotator can assign labels to data points. It may also contain algorithm modules (e.g., active learning) to assist the annotation process.
In the following, we list some of the most stared open source data labeling tools on github as examples:
Tool | Supported Data Type & Label Task |
---|---|
LabelImg | image bounding box |
LabelMe (python) | image segmentation |
Label Studio | multiple |
CVAT) | image & video segmentation |
doccano | text |
VoTT | image & video object detection |