[ Corina Iovan dans la base Horizon ] @incollection{fdi:010080825, title = {{L}'intelligence artificielle au secours de la biodiversit{\'e}}, author = {{M}angeas, {M}organ and {I}ovan, {C}orina and {V}igliola, {L}aurent}, language = {{FRE}}, keywords = {{GESTION} {DE} {L}'{ENVIRONNEMENT} ; {CONSERVATION} {DE} {LA} {NATURE} ; {DIVERSITE} {SPECIFIQUE} ; {COLLECTE} {DE} {DONNEES} ; {INTELLIGENCE} {ARTIFICIELLE} ; {BIODIVERSITE} ; {RECHERCHE} {PLURIDISCIPLINAIRE} ; {FOUILLE} {DE} {DONNEES} ; {ENTREPOT} {DE} {DONNEES} ; {ZONE} {TROPICALE}}, booktitle = {{B}iodiversit{\'e} au {S}ud : recherches pour un monde durable}, pages = {20-21}, adress = {{M}arseille}, publisher = {{IRD}}, year = {2020}, ISBN = {978-2-7099-2850-2}, URL = {http://www.documentation.ird.fr/hor/fdi:010080825}, } @article{fdi:010073616, title = {{C}hange detection of {B}unaken {I}sland coral reefs using 15 years of very high resolution satellite images : a kaleidoscope of habitat trajectories}, author = {{A}mpou, {E}. {E}. and {O}uillon, {S}ylvain and {I}ovan, {C}orina and {A}ndr{\'e}fou{\¨e}t, {S}erge}, language = {{ENG}}, abstract = {{I}n {B}unaken {I}sland ({I}ndonesia), a time-series of very high resolution (2-4 m) satellite imagery was used to draw the long-term dynamics of shallow reef flat habitats from 2001 to 2015. {L}ack of historical georeferenced ground-truth data oriented the analysis towards a scenario-approach based on the monitoring of selected unambiguously-changing habitat polygons characterized in situ in 2014 and 2015. {E}ight representative scenarios (coral colonization, coral loss, coral stability, and sand colonization by seagrass) were identified. {A}ll occurred simultaneously in close vicinity, precluding the identification of a single general cause of changes that could have affected the whole reef. {L}ikely, very fine differences in reef topography, exposure to wind/wave and sea level variations were responsible for the variety of trajectories. {W}hile trajectories of reef habitats is a way to measure resilience and coral recovery, here, the 15-year time-series was too short to be able to conclude on the resilience of {B}unaken reefs.}, keywords = {{I}ndonesia ; {INDESO} ; {R}emote sensing ; {R}eef flat ; {M}arine habitat ; {R}esilience ; {INDONESIE} ; {BUNAKEN} {ILE}}, booktitle = {{I}ndonesia seas management}, journal = {{M}arine {P}ollution {B}ulletin}, volume = {131 {P}art {B}}, numero = {{N}o sp{\'e}cial}, pages = {83--95}, ISSN = {0025-326{X}}, year = {2018}, DOI = {10.1016/j.marpolbul.2017.10.067}, URL = {http://www.documentation.ird.fr/hor/fdi:010073616}, } @article{fdi:010074154, title = {{F}inding the right fit : comparative cetacean distribution models using multiple data sources and statistical approaches}, author = {{D}erville, {S}. and {T}orres, {L}. {G}. and {I}ovan, {C}orina and {G}arrigue, {C}laire}, language = {{ENG}}, abstract = {{A}im: {A}ccurate predictions of cetacean distributions are essential to their conservation but are limited by statistical challenges and a paucity of data. {T}his study aimed at comparing the capacity of various statistical algorithms to deal with biases commonly found in nonsystematic cetacean surveys and to evaluate the potential for citizen science data to improve habitat modelling and predictions. {A}n endangered population of humpback whales ({M}egaptera novaeangliae) in their breeding ground was used as a case study. {L}ocation: {N}ew {C}aledonia, {O}ceania. {M}ethods: {F}ive statistical algorithms were used to model the habitat preferences of humpback whales from 1,360 sightings collected over 14 years of nonsystematic research surveys. {T}hree different background sampling approaches were tested when developing models from 625 crowdsourced sightings to assess methods accounting for citizen science spatial sampling bias. {M}odel evaluation was conducted through cross-validation and prediction to an independent satellite tracking dataset. {R}esults: {A}lgorithms differed in complexity of the environmental relationships modelled, ecological interpretability and transferability. {W}hile parameter tuning had a great effect on model performances, {GLM}s generally had low predictive performance, {SVM}s were particularly hard to interpret, and {BRT}s had high descriptive power but showed signs of overfitting. {MAXENT} and especially {GAM}s provided a valuable complexity trade-off, accurate predictions and were ecologically intelligible. {M}odels showed that humpback whales favoured cool (22-23 degrees {C}) and shallow waters (0-100 m deep) in coastal as well as offshore areas. {C}itizen science models converged with research survey models, specifically when accounting for spatial sampling bias. {M}ain conclusions: {M}arine megafauna distribution models present specific challenges that may be addressed through integrative evaluation, independent testing and appropriately tuned statistical algorithms. {S}pecifically, controlling overfitting is a priority when predicting cetacean distributions for large-scale conservation perspectives. {C}itizen science data appear to be a powerful tool to describe cetacean habitat.}, keywords = {citizen science ; generalized regression ; humpback whales ; machine learning ; species distribution models ; support vector machines ; {NOUVELLE} {CALEDONIE} ; {PACIFIQUE}}, journal = {{D}iversity and {D}istributions}, volume = {24}, numero = {11}, pages = {1657--1673}, ISSN = {1366-9516}, year = {2018}, DOI = {10.1111/ddi.12782}, URL = {http://www.documentation.ird.fr/hor/fdi:010074154}, } @incollection{fdi:010067244, title = {{C}hange detection of coral reef habitats from multi-temporal and multi-source satellite imagery in {B}unaken, {I}ndonesia}, author = {{I}ovan, {C}orina and {A}mpou, {E}. and {A}ndrefou{\¨e}t, {S}erge and {O}uillon, {S}ylvain and {G}aspar, {P}.}, language = {{ENG}}, abstract = {{T}his paper presents a framework for change detection of coral reef habitats in muti-temporal and multi-spectral satellite imagery acquired by multiple sensors. {O}ur specific goal is to analyze the evolution coral reef habitats from images acquired over a period of twelve years. {O}ur study area is located in the centre of the {C}oral {T}riangle, a hotspot of biodiversity in the {B}unaken {I}sland, {I}ndonesia. {I}n-situ data is used in a supervised classification approach to build classification models for eleven habitat types. {R}adiometric calibration is pairwise performed between each image in the data set and the reference one, and habitat models built on the reference image are used to classify the entire time series. {R}esults obtained are discussed and the influential factors are put forward.}, keywords = {{INDONESIE} ; {MER} {DE} {CORAIL} ; {BUNAKEN} {ILE}}, booktitle = {{M}ultitemp 2015}, pages = {4 p.}, publisher = {{IEEE}}, year = {2015}, DOI = {10.1109/{M}ulti-{T}emp.2015.7245758}, ISBN = {978-1-4673-7119-3}, URL = {http://www.documentation.ird.fr/hor/fdi:010067244}, } @incollection{fdi:010067245, title = {{I}ndividual tree segmentation over large areas using airborne {L}i{DAR} point cloud and very high resolution optical imagery}, author = {{Q}in, {Y}. and {F}erraz, {A}. and {M}allet, {C}. and {I}ovan, {C}orina}, language = {{ENG}}, abstract = {{T}imely and accurate measuremen ts of forest parameters are critical for ecosystem studies, sustainable forest resources management, monitoring and planning. {T}his paper presents a processing chain for individual tree segmentation over large areas with airborne {L}i{DAR} 3{D} point cloud and very high resolution ({VHR}) optical imagery. {T}he proposed processing chain consists of fo rest stand level delineation with optical imagery, individual tree segmentation with {C}anopy {H}eight {M}odel ({CHM}) derived from {L}i{DAR} point cloud, rough characterization of trees at forest stand level, and point clustering of individual tree with an {A}daptive {M}ean {S}hift 3{D} ({AMS}3{D}) algorithm. {T}he processing chain is developed with the expectation of supporting operational forest inventory at individual tree level. {E}xperiment is conducted using {L}i{DAR} data acquired in {V}entoux region, {F}rance. {R}esults suggest that the proposed processing chain can be successfully adopted for individual tree characterization over large areas with different forest stands.}, keywords = {{FRANCE}}, booktitle = {{M}ultitemp 2015}, pages = {800--803}, publisher = {{IEEE}}, year = {2015}, ISBN = {978-1-4673-7119-3}, URL = {http://www.documentation.ird.fr/hor/fdi:010067245}, }