We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CV

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computer Vision and Pattern Recognition

Title: How Low Can We Go? Pixel Annotation for Semantic Segmentation

Abstract: How many labeled pixels are needed to segment an image, without any prior knowledge? We conduct an experiment to answer this question.
In our experiment, an Oracle is using Active Learning to train a network from scratch. The Oracle has access to the entire label map of the image, but the goal is to reveal as little pixel labels to the network as possible. We find that, on average, the Oracle needs to reveal (i.e., annotate) less than 0.1% of the pixels in order to train a network. The network can then label all pixels in the image at an accuracy of more than 98%.
Based on this single-image-annotation experiment, we design an experiment to quickly annotate an entire data set. In the data set level experiment the Oracle trains a new network for each image from scratch. The network can then be used to create pseudo-labels, which are the network predicted labels of the unlabeled pixels, for the entire image. Only then, a data set level network is trained from scratch on all the pseudo-labeled images at once.
We repeat both image level and data set level experiments on two, very different, real-world data sets, and find that it is possible to reach the performance of a fully annotated data set using a fraction of the annotation cost.
Comments: Paper and Supplementary
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Cite as: arXiv:2201.10448 [cs.CV]
  (or arXiv:2201.10448v2 [cs.CV] for this version)

Submission history

From: Daniel Kigli [view email]
[v1] Tue, 25 Jan 2022 16:50:24 GMT (17956kb,D)
[v2] Sun, 13 Feb 2022 19:41:52 GMT (17946kb,D)

Link back to: arXiv, form interface, contact.