Imagegeolocalizationisreceivingincreasingattention duetoitsimportanceinseveralapplications,suchasimageretrieval, criminalinvestigationsandfact-checking.Previousworksfocused onseveralinstancesofimagegeolocalizationincludingplacerecognition,GPScoordinatesestimationandcountryrecognition.Inthis paper,wetackleanevenmorechallengingproblem,whichisrecognizingthecitywhereanimagehasbeentaken.Duetothevast numberofcitiesintheworld,wecasttheproblemasaverification problem,wherebythesystemhastodecidewhetheracertainimage hasbeentakeninagivencityornot.Inparticular,wepresentasystemthatgivenaqueryimageandasmallsetofimagestakenina targetcity,decidesifthequeryimagehasbeenshotinthetargetcity ornot.Toallowthesystemtohandlethecaseofimages,takenin citiesthathavenotbeenusedduringtraining,weuseaSiamesenetworkbasedonVisionTransformerasabackbone.Theexperiments werunprovethevalidityoftheproposedsystemwhichoutperforms solutionsbasedonstate-of-the-arttechniques,eveninthechallengingcaseofimagesshotindifferentcitiesofthesamecountry.

Alamayreh, O., Wang, J., Dimitri, G.M., Tondi, B., Barni, M. (2023). A Siamese Based System for City Verification. In ECAI 2023 (pp.69-76) [10.3233/FAIA230255].

A Siamese Based System for City Verification

Alamayreh, Omran
;
Wang, Jun;Dimitri, Giovanna Maria;Tondi, Benedetta;Barni, Mauro
2023-01-01

Abstract

Imagegeolocalizationisreceivingincreasingattention duetoitsimportanceinseveralapplications,suchasimageretrieval, criminalinvestigationsandfact-checking.Previousworksfocused onseveralinstancesofimagegeolocalizationincludingplacerecognition,GPScoordinatesestimationandcountryrecognition.Inthis paper,wetackleanevenmorechallengingproblem,whichisrecognizingthecitywhereanimagehasbeentaken.Duetothevast numberofcitiesintheworld,wecasttheproblemasaverification problem,wherebythesystemhastodecidewhetheracertainimage hasbeentakeninagivencityornot.Inparticular,wepresentasystemthatgivenaqueryimageandasmallsetofimagestakenina targetcity,decidesifthequeryimagehasbeenshotinthetargetcity ornot.Toallowthesystemtohandlethecaseofimages,takenin citiesthathavenotbeenusedduringtraining,weuseaSiamesenetworkbasedonVisionTransformerasabackbone.Theexperiments werunprovethevalidityoftheproposedsystemwhichoutperforms solutionsbasedonstate-of-the-arttechniques,eveninthechallengingcaseofimagesshotindifferentcitiesofthesamecountry.
2023
9781643684369
9781643684376
Alamayreh, O., Wang, J., Dimitri, G.M., Tondi, B., Barni, M. (2023). A Siamese Based System for City Verification. In ECAI 2023 (pp.69-76) [10.3233/FAIA230255].
File in questo prodotto:
File Dimensione Formato  
FAIA-372-FAIA230255.pdf

accesso aperto

Tipologia: PDF editoriale
Licenza: Creative commons
Dimensione 502.65 kB
Formato Adobe PDF
502.65 kB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11365/1250254