Show simple item record

Vision Transformers, a new approach for high-resolution and large-scale mapping of canopy heights

hal.structure.identifierKayrros
hal.structure.identifierLaboratoire des Sciences du Climat et de l'Environnement [Gif-sur-Yvette] [LSCE]
hal.structure.identifierModélisation des Surfaces et Interfaces Continentales [MOSAIC]
dc.contributor.authorFAYAD, Ibrahim
hal.structure.identifierLaboratoire des Sciences du Climat et de l'Environnement [Gif-sur-Yvette] [LSCE]
dc.contributor.authorCIAIS, Philippe
hal.structure.identifierLaboratoire des Sciences du Climat et de l'Environnement [Gif-sur-Yvette] [LSCE]
hal.structure.identifierModélisation des Surfaces et Interfaces Continentales [MOSAIC]
dc.contributor.authorSCHWARTZ, Martin
hal.structure.identifierInteractions Sol Plante Atmosphère [UMR ISPA]
dc.contributor.authorWIGNERON, Jean-Pierre
hal.structure.identifierTerritoires, Environnement, Télédétection et Information Spatiale [UMR TETIS]
dc.contributor.authorBAGHDADI, Nicolas
hal.structure.identifierKayrros
dc.contributor.authorDE TRUCHIS, Aurélien
hal.structure.identifierLaboratoire d'informatique de l'école normale supérieure [LIENS]
hal.structure.identifierCentre National de la Recherche Scientifique [CNRS]
hal.structure.identifierStatistical Machine Learning and Parsimony [SIERRA]
hal.structure.identifierUniversité Paris Sciences et Lettres [PSL]
dc.contributor.authorD'ASPREMONT, Alexandre
hal.structure.identifierInteractions Sol Plante Atmosphère [UMR ISPA]
dc.contributor.authorFRAPPART, Frederic
hal.structure.identifierCalifornia Institute of Technology [CALTECH]
dc.contributor.authorSAATCHI, Sassan
hal.structure.identifierProcessus d'Activation Sélective par Transfert d'Energie Uni-électronique ou Radiatif (UMR 8640) [PASTEUR]
dc.contributor.authorPELLISSIER-TANON, Agnes
hal.structure.identifierAtos
dc.contributor.authorBAZZI, Hassan
dc.date.accessioned2024-04-08T11:42:30Z
dc.date.available2024-04-08T11:42:30Z
dc.date.issued2023-04-22
dc.identifier.urihttps://oskar-bordeaux.fr/handle/20.500.12278/195072
dc.description.abstractEnAccurate and timely monitoring of forest canopy heights is critical for assessing forest dynamics, biodiversity, carbon sequestration as well as forest degradation and deforestation. Recent advances in deep learning techniques, coupled with the vast amount of spaceborne remote sensing data offer an unprecedented opportunity to map canopy height at high spatial and temporal resolutions. Current techniques for wall-to-wall canopy height mapping correlate remotely sensed 2D information from optical and radar sensors to the vertical structure of trees using LiDAR measurements. While studies using deep learning algorithms have shown promising performances for the accurate mapping of canopy heights, they have limitations due to the type of architectures and loss functions employed. Moreover, mapping canopy heights over tropical forests remains poorly studied, and the accurate height estimation of tall canopies is a challenge due to signal saturation from optical and radar sensors, persistent cloud covers and sometimes the limited penetration capabilities of LiDARs. Here, we map heights at 10 m resolution across the diverse landscape of Ghana with a new vision transformer (ViT) model optimized concurrently with a classification (discrete) and a regression (continuous) loss function. This model achieves better accuracy than previously used convolutional based approaches (ConvNets) optimized with only a continuous loss function. The ViT model results show that our proposed discrete/continuous loss significantly increases the sensitivity for very tall trees (i.e., > 35m), for which other approaches show saturation effects. The height maps generated by the ViT also have better ground sampling distance and better sensitivity to sparse vegetation in comparison to a convolutional model. Our ViT model has a RMSE of 3.12m in comparison to a reference dataset while the ConvNet model has a RMSE of 4.3m.
dc.description.sponsorshipPaRis Artificial Intelligence Research InstitutE - ANR-19-P3IA-0001
dc.language.isoen
dc.subject.encanopy height
dc.subject.enGEDI
dc.subject.enSentinel 1
dc.subject.enSentinel 2
dc.subject.enVision Transformers
dc.subject.enDeep learning
dc.subject.enKnowledge distillation
dc.title.enVision Transformers, a new approach for high-resolution and large-scale mapping of canopy heights
dc.typeDocument de travail - Pré-publication
dc.identifier.doi10.48550/arXiv.2304.11487
dc.subject.halMathématiques [math]
dc.identifier.arxiv2304.11487
bordeaux.hal.laboratoriesInteractions Soil Plant Atmosphere (ISPA) - UMR 1391*
bordeaux.institutionBordeaux Sciences Agro
bordeaux.institutionINRAE
hal.identifierhal-04230906
hal.version1
hal.origin.linkhttps://hal.archives-ouvertes.fr//hal-04230906v1
bordeaux.COinSctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.date=2023-04-22&rft.au=FAYAD,%20Ibrahim&CIAIS,%20Philippe&SCHWARTZ,%20Martin&WIGNERON,%20Jean-Pierre&BAGHDADI,%20Nicolas&rft.genre=preprint


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record