Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Seed, Expand and Constrain: Three Principles for Weakly-Supervised Image Segmentation

About

We introduce a new loss function for the weakly-supervised training of semantic image segmentation models based on three guiding principles: to seed with weak localization cues, to expand objects based on the information about which classes can occur in an image, and to constrain the segmentations to coincide with object boundaries. We show experimentally that training a deep convolutional neural network using the proposed loss function leads to substantially better segmentations than previous state-of-the-art methods on the challenging PASCAL VOC 2012 dataset. We furthermore give insight into the working mechanism of our method by a detailed experimental study that illustrates how the segmentation quality is affected by each term of the proposed loss function as well as their combinations.

Alexander Kolesnikov, Christoph H. Lampert• 2016

Related benchmarks

TaskDatasetResultRank
Semantic segmentationPASCAL VOC 2012 (val)
Mean IoU50.7
2142
Semantic segmentationPASCAL VOC 2012 (test)
mIoU51.7
1415
Semantic segmentationCamVid (test)
mIoU2.5
411
Semantic segmentationCOCO 2014 (val)
mIoU22.4
304
Weakly supervised semantic segmentationPASCAL VOC 2012 (val)
mIoU50.7
168
Weakly supervised semantic segmentationPASCAL VOC 2012 (test)
mIoU51.7
158
Semantic segmentationCOCO (val)
mIoU22.4
150
Semantic segmentationCOCO Object (val)
mIoU0.224
97
Weakly supervised semantic segmentationMS-COCO 2014 (val)
mIoU22.4
27
Weakly supervised semantic segmentationVOC 2012 (val)
mIoU50.7
19
Showing 10 of 10 rows

Other info

Follow for update