Panoptic SwiftNet: Pyramidal Fusion for Real-Time Panoptic Segmentation
-
Published:2023-04-07
Issue:8
Volume:15
Page:1968
-
ISSN:2072-4292
-
Container-title:Remote Sensing
-
language:en
-
Short-container-title:Remote Sensing
Author:
Šarić Josip1, Oršić Marin2, Šegvić Siniša1ORCID
Affiliation:
1. Faculty of Electrical Engineering and Computing, University of Zagreb, 10000 Zagreb, Croatia 2. Microblink, 10000 Zagreb, Croatia
Abstract
Dense panoptic prediction is a key ingredient in many existing applications such as autonomous driving, automated warehouses, or remote sensing. Many of these applications require fast inference over large input resolutions on affordable or even embedded hardware. We proposed to achieve this goal by trading off backbone capacity for multi-scale feature extraction. In comparison with contemporaneous approaches to panoptic segmentation, the main novelties of our method are efficient scale-equivariant feature extraction, cross-scale upsampling through pyramidal fusion and boundary-aware learning of pixel-to-instance assignment. The proposed method is very well suited for remote sensing imagery due to the huge number of pixels in typical city-wide and region-wide datasets. We present panoptic experiments on Cityscapes, Vistas, COCO, and the BSB-Aerial dataset. Our models outperformed the state-of-the-art on the BSB-Aerial dataset while being able to process more than a hundred 1MPx images per second on an RTX3090 GPU with FP16 precision and TensorRT optimization.
Funder
Rimac Technology Croatian Science Foundation European Regional Development Fund
Subject
General Earth and Planetary Sciences
Reference47 articles.
1. Kirillov, A., He, K., Girshick, R., Rother, C., and Dollár, P. (2019, January 15–20). Panoptic segmentation. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, CA, USA. 2. Faster r-cnn: Towards real-time object detection with region proposal networks;Ren;Adv. Neural Inf. Process. Syst.,2015 3. Long, J., Shelhamer, E., and Darrell, T. (2015, January 7–12). Fully convolutional networks for semantic segmentation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Boston, MA, USA. 4. Zendel, O., Schörghuber, M., Rainer, B., Murschitz, M., and Beleznai, C. (2022, January 18–24). Unifying panoptic segmentation for autonomous driving. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, LA, USA. 5. Garnot, V.S.F., and Landrieu, L. (2021, January 11–17). Panoptic segmentation of satellite image time series with convolutional temporal attention networks. Proceedings of the IEEE/CVF International Conference on Computer Vision, Montreal, BC, Canada.
Cited by
3 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
|
|