From Global Image Annotation to Interactive Object Segmentation
| Resource Type | Date |
|---|---|
| Results | 2012-08-01 |
Description
Results related to the submission to the Special Issue on “Methods and Tools for Ground Truth Collection in Multimedia Applications” of the Multimedia Tools and Applications (MTAP) Journal, Springer.
- Description of the used dataset.
- Bounding boxes: Masks generated with the six modes of region selection and the two upper bounds.
- Scribbles: Masks generated with the three modes of label propagation.
- User tests: Masks and time totals, averaged by user and averaged by image.
These results have been produced by Xavier Giró-i-Nieto, Manuel Martos, Eva Mohedano and Jordi Pont-Tuset.
People involved
| Jordi Pont-Tuset | PhD Candidate |
| Xavier Giró | Associate Professor |
Related Publications
, “From Global Image Annotation to Interactive Object Segmentation”, Multimedia Tools and Applications, vol. 70, 2014. (1.71 MB) |
|
, “Image Segmentation Evaluation and Its Application to Object Detection”, Universitat Politècnica de Catalunya, BarcelonaTech, Barcelona, 2014. (48.44 MB) |