Pix2Point : learning outdoor 3D using sparse point clouds and optimal transport
| Autor: | B. Le Saux, Marcela Carvalho, Frédéric Champagnat, Rémy Leroy, Pauline Trouvé-Peloux |
|---|---|
| Přispěvatelé: | DTIS, ONERA, Université Paris Saclay [Palaiseau], ONERA-Université Paris-Saclay, ESA Centre for Earth Observation (ESRIN), European Space Agency (ESA), UPCITY, Montreuil, France, GREC, christine |
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Předmět: |
FOS: Computer and information sciences
Minimisation (psychology) 0209 industrial biotechnology ESTIMATION 3D NUAGES DE POINTS [SPI] Engineering Sciences [physics] Machine vision Computer science Computer Vision and Pattern Recognition (cs.CV) Point cloud Computer Science - Computer Vision and Pattern Recognition ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION [MATH] Mathematics [math] 02 engineering and technology [INFO] Computer Science [cs] [PHYS] Physics [physics] Hybrid neural network [SPI]Engineering Sciences [physics] 020901 industrial engineering & automation 0202 electrical engineering electronic engineering information engineering Computer vision [INFO]Computer Science [cs] [MATH]Mathematics [math] Divergence (statistics) [PHYS]Physics [physics] I.2.10 Monocular Artificial neural network I.4.5 business.industry Deep learning APPRENTISSAGE PROFOND 020201 artificial intelligence & image processing Artificial intelligence business |
| Zdroj: | MVA 2021 MVA 2021, Jul 2021, VIRTUEL, Japan HAL MVA |
| Popis: | Good quality reconstruction and comprehension of a scene rely on 3D estimation methods. The 3D information was usually obtained from images by stereo-photogrammetry, but deep learning has recently provided us with excellent results for monocular depth estimation. Building up a sufficiently large and rich training dataset to achieve these results requires onerous processing. In this paper, we address the problem of learning outdoor 3D point cloud from monocular data using a sparse ground-truth dataset. We propose Pix2Point, a deep learning-based approach for monocular 3D point cloud prediction, able to deal with complete and challenging outdoor scenes. Our method relies on a 2D-3D hybrid neural network architecture, and a supervised end-to-end minimisation of an optimal transport divergence between point clouds. We show that, when trained on sparse point clouds, our simple promising approach achieves a better coverage of 3D outdoor scenes than efficient monocular depth methods. 5 pages, 2 figures, to be published in 2021 International Conference on Machine Vision Applications |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|