System and method for automated content labeling
An automated content labeling system is disclosed. An example embodiment is configured to: register a plurality of labelers to which annotation tasks are assigned; populate a labeling queue with content data to be annotated; assign annotation tasks from the labeling queue to the plurality of labelers; and provide a superpixel annotation tool enabling the plurality of labelers to configure a size of a segment cluster in an image of the content data, and select each segment cluster to be included in a segmentation feature with a specified object class, the segment clusters including similarly colored pixels from the image.
1. An automated content labeling system, the system comprising:
a data processor; and
an automated content labeling platform, executed by the data processor, the automated content labeling platform being configured to:
register a plurality of labelers to which annotation tasks are assigned;
populate, by use of the data processor, a labeling queue with content data to be annotated;
assign, by use of the data processor, annotation tasks from the labeling queue to the plurality of labelers, the annotation tasks having associated datasets representing sets of content data to be annotated by the plurality of labelers;
prompt each of the plurality of labelers to begin processing through the datasets and apply labels to objects identified in the content data;
provide, by use of the data processor, a segmentation tool enabling the plurality of labelers to configure a size of a segment cluster in an image of the content data, and select each segment cluster to be included in a segmentation feature with a specified object class, the object class corresponding to an object label for the segmentation feature, the segment clusters including similarly colored pixels from the image; and
generate, by the use of the data processor, an auto consensus score corresponding to a level of conformity of a label applied to a particular item of the content data by a particular labeler of the plurality of labelers with other labels applied to the particular item of the content data by others of the plurality of labelers, the auto consensus score is displayed to the particular labeler.
2. The automated content labeling system of claim 1 being further configured to provide an electronic pen tool and an electronic eraser tool enabling a labeler of the plurality of labelers to adjust boundaries of the segmentation feature.
3. The automated content labeling system of claim 1 being further configured to provide a tool to overwrite the segmentation feature.
4. The automated content labeling system of claim 1 being further configured to enable a labeler of the plurality of labelers to create multiple instances of object annotations with a same class designation.
5. The automated content labeling system of claim 1 being further configured to enable a labeler of the plurality of labelers to create a nested classification of an object.
6. The automated content labeling system of claim 1 being further configured to enable a labeler of the plurality of labelers to identify an object in the content data using a multi-frame bounding box.
7. The automated content labeling system of claim 6 being further configured to assign a keyframe to the multi-frame bounding box.
8. The automated content labeling system of claim 1 being further configured to enable a labeler of the plurality of labelers to label a text string in the content data.
9. The automated content labeling system of claim 1 wherein the content data is of a type from the group consisting of: images, textual content, numerical content, audio data, chemical signatures, and organic signatures.
10. A method comprising:
registering, by use of a data processor, a plurality of labelers to which annotation tasks are assigned;
populating, by use of the data processor, a labeling queue with content data to be annotated;
assigning, by use of the data processor, annotation tasks from the labeling queue to the plurality of labelers, the annotation tasks having associated datasets representing sets of content data to be annotated by the plurality of labelers;
prompting each of the plurality of labelers to begin processing through the datasets and apply labels to objects identified in the content data;
providing, by use of the data processor, a segmentation tool enabling the plurality of labelers to configure a size of a segment cluster in an image of the content data, and select each segment cluster to be included in a segmentation feature with a specified object class, the object class corresponding to an object label for the segmentation feature, the segment clusters including similarly colored pixels from the image; and
generate, by the use of the data processor, an auto consensus score corresponding to a level of conformity of a label applied to a particular item of the content data by a particular labeler of the plurality of labelers with other labels applied to the particular item of the content data by others of the plurality of labelers, the auto consensus score is displayed to the particular labeler.
11. The method of claim 10 including providing an electronic pen tool and an electronic eraser tool enabling a labeler of the plurality of labelers to adjust boundaries of the segmentation feature.
12. The method of claim 10 including providing a tool to overwrite the segmentation feature.
13. The method of claim 10 including enabling a labeler of the plurality of labelers to create multiple instances of object annotations with a same class designation.
14. The method of claim 10 including enabling a labeler of the plurality of labelers to create a nested classification of an object.
15. The method of claim 10 including enabling a labeler of the plurality of labelers to identify an object in the content data using a multi-frame bounding box.
16. The method of claim 15 including assigning a keyframe to the multi-frame bounding box.
17. The method of claim 10 including enabling a labeler of the plurality of labelers to label a text string in the content data.
18. The method of claim 10 wherein the content data is of a type from the group consisting of: images, textual content, numerical content, audio data, chemical signatures, and organic signatures.
19. A non-transitory machine-readable storage medium embodying instructions which, when executed by a processor, cause the processor to:
register, by use of a data processor, a plurality of labelers to which annotation tasks are assigned;
populate, by use of the data processor, a labeling queue with content data to be annotated;
assign, by use of the data processor, annotation tasks from the labeling queue to the plurality of labelers, the annotation tasks having associated datasets representing sets of content data to be annotated by the plurality of labelers;
prompt each of the plurality of labelers to begin processing through the datasets and apply labels to objects identified in the content data;
provide, by use of the data processor, a segmentation tool enabling the plurality of labelers to configure a size of a segment cluster in an image of the content data, and select each segment cluster to be included in a segmentation feature with a specified object class, the object class corresponding to an object label for the segmentation feature, the segment clusters including similarly colored pixels from the image; and
generate, by the use of the data processor, an auto consensus score corresponding to a level of conformity of a label applied to a particular item of the content data by a particular labeler of the plurality of labelers with other labels applied to the particular item of the content data by others of the plurality of labelers, the auto consensus score is displayed to the particular labeler.
20. The non-transitory machine-readable storage medium embodying the instructions of claim 19 wherein the content data is of a type from the group consisting of: images, textual content, numerical content, audio data, chemical signatures, and organic signatures.