Abstract | ||
---|---|---|
Semantic instance segmentation remains a challenge. We propose to tackle the problem with a discriminative loss function, operating at pixel level, that encourages a convolutional network to produce a representation of the image that can easily be clustered into instances with a simple post-processing step. Our approach of combining an off-the-shelf network with a principled loss function inspired by a metric learning objective is conceptually simple and distinct from recent efforts in instance segmentation and is well-suited for real-time applications. In contrast to previous works, our method does not rely on object proposals or recurrent mechanisms and is particularly well suited for tasks with complex occlusions. A key contribution of our work is to demonstrate that such a simple setup without bells and whistles is effective and can perform on-par with more complex methods. We achieve competitive performance on the Cityscapes segmentation benchmark. |
Year | DOI | Venue |
---|---|---|
2017 | 10.1109/CVPRW.2017.66 | 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) |
Keywords | Field | DocType |
semantic instance segmentation,autonomous driving,off-the-shelf network,metric learning objective,principled loss function,complex occlusions,Cityscapes segmentation benchmark | Scale-space segmentation,Computer science,Segmentation-based object categorization,Image segmentation,Artificial intelligence,Discriminative model,Benchmark (computing),Computer vision,Pattern recognition,Segmentation,Pixel,Semantics,Machine learning | Conference |
Volume | Issue | ISSN |
2017 | 1 | 2160-7508 |
ISBN | Citations | PageRank |
978-1-5386-0734-3 | 5 | 0.42 |
References | Authors | |
13 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Bert De Brabandere | 1 | 8 | 1.56 |
Davy Neven | 2 | 32 | 2.11 |
Luc Van Gool | 3 | 27566 | 1819.51 |