Scipediacontent: Scipediacontent moved page Draft Content 337910366 to Ferrera et al 2019a

2021-02-02T05:49:59Z

Scipediacontent moved page Draft Content 337910366 to Ferrera et al 2019a

Scipediacontent: Created page with " == Abstract == International audience; The estimation of disparity maps from stereo pairs has many applications in robotics and autonomous driving. Stereo matching has first..."

2021-02-02T05:49:53Z

Created page with " == Abstract == International audience; The estimation of disparity maps from stereo pairs has many applications in robotics and autonomous driving. Stereo matching has first..."

New page

== Abstract ==

International audience; The estimation of disparity maps from stereo pairs has many applications in robotics and autonomous driving. Stereo matching has first been solved using model-based approaches, with real-time considerations for some, but to-day's most recent works rely on deep convolutional neural networks and mainly focus on accuracy at the expense of computing time. In this paper, we present a new method for disparity maps estimation getting the best of both worlds: the accuracy of data-based methods and the speed of fast model-based ones. The proposed approach fuses prior disparity maps to estimate a refined version. The core of this fusion pipeline is a convolutional neural network that leverages dilated convolutions for fast context aggregation without spatial resolution loss. The resulting architecture is both very effective for the task of refining and fusing prior disparity maps and very light, allowing our fusion pipeline to produce disparity maps at rates up to 125 Hz. We obtain state-of-the-art results in terms of speed and accuracy on the KITTI benchmarks. Code and pre-trained models are available on our github: https://github.com/ ferreram/FD-Fusion.

== Original document ==

The different versions of the original document can be found in:

* [https://hal.archives-ouvertes.fr/hal-02326896/file/_3DV2019__Deep_Refinement_CameraReady_IEEE_validated.pdf https://hal.archives-ouvertes.fr/hal-02326896/file/_3DV2019__Deep_Refinement_CameraReady_IEEE_validated.pdf]

* [http://xplorestaging.ieee.org/ielx7/8882741/8885410/08886031.pdf?arnumber=8886031 http://xplorestaging.ieee.org/ielx7/8882741/8885410/08886031.pdf?arnumber=8886031],
: [http://dx.doi.org/10.1109/3dv.2019.00011 http://dx.doi.org/10.1109/3dv.2019.00011]

* [https://hal.archives-ouvertes.fr/hal-02326896 https://hal.archives-ouvertes.fr/hal-02326896],
: [https://dblp.uni-trier.de/db/conf/3dim/3dim2019.html#FerreraBM19 https://dblp.uni-trier.de/db/conf/3dim/3dim2019.html#FerreraBM19],
: [https://ieeexplore.ieee.org/abstract/document/8886031 https://ieeexplore.ieee.org/abstract/document/8886031],
: [https://doi.org/10.1109/3DV.2019.00011 https://doi.org/10.1109/3DV.2019.00011],
: [https://academic.microsoft.com/#/detail/2982435595 https://academic.microsoft.com/#/detail/2982435595]

* [https://hal.archives-ouvertes.fr/hal-02326896 https://hal.archives-ouvertes.fr/hal-02326896],
: [https://hal.archives-ouvertes.fr/hal-02326896/document https://hal.archives-ouvertes.fr/hal-02326896/document],
: [https://hal.archives-ouvertes.fr/hal-02326896/file/_3DV2019__Deep_Refinement_CameraReady_IEEE_validated.pdf https://hal.archives-ouvertes.fr/hal-02326896/file/_3DV2019__Deep_Refinement_CameraReady_IEEE_validated.pdf]

← Older revision	Revision as of 05:49, 2 February 2021
(No difference)

Ferrera et al 2019a - Revision history

Scipediacontent: Scipediacontent moved page Draft Content 337910366 to Ferrera et al 2019a

Scipediacontent: Created page with " == Abstract == International audience; The estimation of disparity maps from stereo pairs has many applications in robotics and autonomous driving. Stereo matching has first..."