Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2403.03296 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914898493046784 |
|---|---|
| author | Litto, Katia Jodogne-Del Bilodeau, Guillaume-Alexandre |
| author_facet | Litto, Katia Jodogne-Del Bilodeau, Guillaume-Alexandre |
| contents | Increasing the accuracy of instance segmentation methods is often done at the expense of speed. Using coarser representations, we can reduce the number of parameters and thus obtain real-time masks. In this paper, we take inspiration from the set cover problem to predict mask approximations. Given ground-truth binary masks of objects of interest as training input, our method learns to predict the approximate coverage of these objects by disks without supervision on their location or radius. Each object is represented by a fixed number of disks with different radii. In the learning phase, we consider the radius as proportional to a standard deviation in order to compute the error to propagate on a set of two-dimensional Gaussian functions rather than disks. We trained and tested our instance segmentation method on challenging datasets showing dense urban settings with various road users. Our method achieve state-of-the art results on the IDD and KITTI dataset with an inference time of 0.040 s on a single RTX 3090 GPU. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_03296 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | CenterDisks: Real-time instance segmentation with disk covering Litto, Katia Jodogne-Del Bilodeau, Guillaume-Alexandre Computer Vision and Pattern Recognition Increasing the accuracy of instance segmentation methods is often done at the expense of speed. Using coarser representations, we can reduce the number of parameters and thus obtain real-time masks. In this paper, we take inspiration from the set cover problem to predict mask approximations. Given ground-truth binary masks of objects of interest as training input, our method learns to predict the approximate coverage of these objects by disks without supervision on their location or radius. Each object is represented by a fixed number of disks with different radii. In the learning phase, we consider the radius as proportional to a standard deviation in order to compute the error to propagate on a set of two-dimensional Gaussian functions rather than disks. We trained and tested our instance segmentation method on challenging datasets showing dense urban settings with various road users. Our method achieve state-of-the art results on the IDD and KITTI dataset with an inference time of 0.040 s on a single RTX 3090 GPU. |
| title | CenterDisks: Real-time instance segmentation with disk covering |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2403.03296 |