@incollection{fdi:010090482,
  title = {{C}omparing spatial and spatio-temporal paradigms to estimate the evolution of socio-economical indicators from satellite images},
  author = {{J}arry, {R}. and {C}haumont, {M}. and {B}erti-{E}quille, {L}aure and {S}ubsol, {G}.},
  editor = {},
  language = {{ENG}},
  abstract = {{I}n remote sensing, deep spatio-temporal models, i.e., deep learning models that estimate information based on {S}atellite {I}mage {T}ime {S}eries obtain successful results in {L}and {U}se/{L}and {C}over classification or change detection. {N}evertheless, for socioeconomic applications such as poverty estimation, only deep spatial models have been proposed. {I}n this paper, we propose a test-bed to compare spatial and spatio-temporal paradigms to estimate the evolution of {N}ighttime {L}ight ({NTL}), a standard proxy for socioeconomic indicators. {W}e applied the test-bed in the area of {Z}anzibar, {T}anzania for 21 years. {W}e observe that (1) both models obtain roughly equivalent performances when predicting the {NTL} value at a given time, but (2) the spatio-temporal model is significantly more efficient when predicting the {NTL} evolution.},
  keywords = {{ZANZIBAR} ; {TANZANIE}},
  booktitle = {{IGARSS} 2023 - 2023 {IEEE} {I}nternational {G}eoscience and {R}emote {S}ensing {S}ymposium : proceedings},
  numero = {},
  pages = {5790--5793},
  address = {{P}iscataway},
  publisher = {{IEEE}},
  series = {},
  year = {2023},
  DOI = {10.1109/{IGARSS}52108.2023.10282306},
  ISBN = {979-8-3503-2010-7},
  URL = {https://www.documentation.ird.fr/hor/fdi:010090482},
}