{"id":12375,"date":"2024-05-31T17:36:49","date_gmt":"2024-05-31T14:36:49","guid":{"rendered":"https:\/\/www.kaspersky.com.tr\/blog\/?p=12375"},"modified":"2024-05-31T17:36:49","modified_gmt":"2024-05-31T14:36:49","slug":"dense-pose-recognition-from-wi-fi-signal","status":"publish","type":"post","link":"https:\/\/www.kaspersky.com.tr\/blog\/dense-pose-recognition-from-wi-fi-signal\/12375\/","title":{"rendered":"Wi-Fi kullanarak bir ki\u015fiyi bulma ve duru\u015f pozisyonunu alg\u0131lama"},"content":{"rendered":"<p>Diyojen g\u00fcnd\u00fcz vakti d\u00fcr\u00fcst bir adam aramak i\u00e7in elinde bir fenerle gezerken yaln\u0131zca optik alg\u0131lama y\u00f6ntemlerine g\u00fcvenmi\u015ftir. Ancak g\u00fcn\u00fcm\u00fczde bilim insanlar\u0131 bu ama\u00e7la Wi-Fi sinyallerinin kullan\u0131lmas\u0131n\u0131 \u00f6nermektedir. Daha spesifik olmak gerekirse, Carnegie Mellon \u00dcniversitesi\u2019ndeki \u00fc\u00e7 ara\u015ft\u0131rmac\u0131 taraf\u0131ndan <a href=\"https:\/\/arxiv.org\/pdf\/2301.00250.pdf\" target=\"_blank\" rel=\"nofollow noopener\">geli\u015ftirilen<\/a> y\u00f6ntem, s\u0131radan bir ev Wi-Fi y\u00f6nlendiricisinden gelen sinyali kullanarak bir ki\u015finin yaln\u0131zca bir odadaki konumunu de\u011fil, ayn\u0131 zamanda duru\u015f pozisyonunu da tespit ediyor.<\/p>\n<p>Neden Wi-Fi? Bunun birka\u00e7 nedeni vard\u0131r. \u0130lk olarak, optik alg\u0131laman\u0131n aksine, radyo sinyalleri karanl\u0131kta m\u00fckemmel \u00e7al\u0131\u015f\u0131r ve mobilya gibi k\u00fc\u00e7\u00fck engeller taraf\u0131ndan engellenmez. \u0130kincisi ucuz olmas\u0131, zira potansiyel olarak bu i\u015fi yapabilecek di\u011fer ara\u00e7lar olan lidar ve radarlar\u0131n pek de ucuz olduklar\u0131 s\u00f6ylenemez. \u00dc\u00e7\u00fcnc\u00fcs\u00fc, Wi-Fi\u2019\u0131n zaten her yerde olmas\u0131d\u0131r, tek yap\u0131lmas\u0131 gereken o a\u011fa eri\u015fip yakalayabilmektir. Peki bu y\u00f6ntem ne kadar etkili ve bunu kullanarak neler yapabilirsiniz? Hadi biraz daha yak\u0131ndan bakal\u0131m.<\/p>\n<h2>DensePose: G\u00f6r\u00fcnt\u00fclerdeki insan pozlar\u0131n\u0131 alg\u0131lamak i\u00e7in bir y\u00f6ntem<\/h2>\n<p>Ba\u015flamak i\u00e7in biraz geriye gitmemiz laz\u0131m. \u00d6ncelikle insan v\u00fccudunun ve genel olarak pozlar\u0131n\u0131n nas\u0131l do\u011fru bir \u015fekilde alg\u0131lanabilece\u011fini anlamam\u0131z gerekir bu da bizi 2018 y\u0131l\u0131nda ba\u015fka bir grup bilim insan\u0131n\u0131n sundu\u011fu <a href=\"https:\/\/arxiv.org\/pdf\/1802.00434.pdf\" target=\"_blank\" rel=\"nofollow noopener\">DensePose adl\u0131 y\u00f6nteme g\u00f6t\u00fcr\u00fcr<\/a>. Bilim adamlar\u0131n\u0131n foto\u011fraflardaki insan pozlar\u0131n\u0131 alg\u0131lamak, yani derinlik i\u00e7in ek veri i\u00e7ermeyen iki boyutlu g\u00f6r\u00fcnt\u00fcler elde etmek i\u00e7in ba\u015far\u0131yla kulland\u0131klar\u0131 bu y\u00f6ntem \u015f\u00f6yle \u00e7al\u0131\u015f\u0131r: \u0130lk olarak, <a href=\"https:\/\/densepose.org\/\" target=\"_blank\" rel=\"nofollow noopener\">DensePose<\/a> modeli g\u00f6r\u00fcnt\u00fclerde insan v\u00fccudu olarak alg\u0131lanan nesneleri arar. Bu nesneler daha sonra her biri belirli bir v\u00fccut b\u00f6l\u00fcm\u00fcne kar\u015f\u0131l\u0131k gelen farkl\u0131 alanlara ayr\u0131larak tek tek analiz edilir. Bu yakla\u015f\u0131m\u0131n sebebi v\u00fccut par\u00e7alar\u0131 birbirinden \u00e7ok farkl\u0131 hareket etmesidir; \u00f6rne\u011fin ba\u015f ve g\u00f6vde, kollar ve bacaklardan \u00e7ok farkl\u0131 davran\u0131r.<\/p>\n<div id=\"attachment_12377\" style=\"width: 2510px\" class=\"wp-caption aligncenter\"><a href=\"https:\/\/media.kasperskydaily.com\/wp-content\/uploads\/sites\/91\/2024\/05\/31171936\/dense-pose-recognition-from-wi-fi-signal-01.jpeg\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-12377\" class=\"size-full wp-image-12377\" src=\"https:\/\/media.kasperskydaily.com\/wp-content\/uploads\/sites\/91\/2024\/05\/31171936\/dense-pose-recognition-from-wi-fi-signal-01.jpeg\" alt=\"DensePose: Foto\u011fraflardaki insan pozlar\u0131n\u0131 alg\u0131lamaya y\u00f6nelik bir y\u00f6ntem \" width=\"2500\" height=\"600\"><\/a><p id=\"caption-attachment-12377\" class=\"wp-caption-text\">DensePose, foto\u011fraflardaki insan bedenlerinin pozlar\u0131n\u0131 do\u011fru bir \u015fekilde alg\u0131layabiliyor ve hatta y\u00fczeylerinin UV haritalar\u0131n\u0131 olu\u015fturabiliyor. <a href=\"https:\/\/arxiv.org\/pdf\/1802.00434.pdf\" target=\"_blank\" rel=\"nofollow noopener\">Kaynak<\/a><\/p><\/div>\n<p>Sonu\u00e7 olarak model, 2D bir g\u00f6r\u00fcnt\u00fcy\u00fc insan v\u00fccudunun 3D y\u00fczeyiyle ili\u015fkilendirmeyi \u00f6\u011frenerek yaln\u0131zca alg\u0131lanan poza kar\u015f\u0131l\u0131k gelen g\u00f6r\u00fcnt\u00fc ek a\u00e7\u0131klamalar\u0131n\u0131 de\u011fil, ayn\u0131 zamanda foto\u011frafta tasvir edilen v\u00fccudun UV haritas\u0131n\u0131 da elde etmi\u015ftir. Bu ikinci \u00f6zellik, g\u00f6r\u00fcnt\u00fcn\u00fcn \u00fczerine ekstra bir doku yerle\u015ftirmeyi de m\u00fcmk\u00fcn k\u0131lar.<\/p>\n<p>En etkileyici olan\u0131 ise; bu tekni\u011fin grup foto\u011fraflar\u0131ndaki birden fazla ki\u015finin pozlar\u0131n\u0131, hatta insanlar\u0131n bir araya topland\u0131\u011f\u0131 ve birbirlerini k\u0131smen engelledi\u011fi kaotik \u201cmezuniyet balosu\u201d foto\u011fraflar\u0131n\u0131 bile do\u011fru bir \u015fekilde alg\u0131layabilmesidir.<\/p>\n<div id=\"attachment_12380\" style=\"width: 2288px\" class=\"wp-caption aligncenter\"><a href=\"https:\/\/media.kasperskydaily.com\/wp-content\/uploads\/sites\/91\/2024\/05\/31172321\/dense-pose-recognition-from-wi-fi-signal-02.jpeg\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-12380\" class=\"size-full wp-image-12380\" src=\"https:\/\/media.kasperskydaily.com\/wp-content\/uploads\/sites\/91\/2024\/05\/31172321\/dense-pose-recognition-from-wi-fi-signal-02.jpeg\" alt=\"DensePose: Foto\u011fraflarda poz alg\u0131lama \u00f6rnekleri \" width=\"2278\" height=\"1471\"><\/a><p id=\"caption-attachment-12380\" class=\"wp-caption-text\">DensePose, grup foto\u011fraflar\u0131ndaki bireysel fig\u00fcrlerin konumlar\u0131n\u0131 do\u011fru bir \u015fekilde alg\u0131l\u0131yor.<a href=\"https:\/\/arxiv.org\/pdf\/1802.00434.pdf\" target=\"_blank\" rel=\"nofollow noopener\">Kaynak<\/a><\/p><\/div>\n<p>Dahas\u0131, makalede sunulan g\u00f6r\u00fcnt\u00fclere ve ara\u015ft\u0131rmac\u0131lar taraf\u0131ndan yay\u0131nlanan videolara inan\u0131lacak olursa, sistem en s\u0131ra d\u0131\u015f\u0131 v\u00fccut pozisyonlar\u0131n\u0131 bile rahatl\u0131kla alg\u0131layabilir. \u00d6rne\u011fin bu yapay sinir a\u011f\u0131; bisiklet, motosiklet ve at s\u0131rt\u0131ndaki insanlar\u0131 do\u011fru bir \u015fekilde alg\u0131layabilir ve ayr\u0131ca beyzbol oyuncular\u0131n\u0131n, futbolcular\u0131n ve hatta genellikle \u00f6ng\u00f6r\u00fclemeyen \u015fekillerde hareket eden break dans\u00e7\u0131lar\u0131n pozlar\u0131n\u0131 dahi do\u011fru bir \u015fekilde belirleyebilir.<\/p>\n<div id=\"attachment_12381\" style=\"width: 2290px\" class=\"wp-caption aligncenter\"><a href=\"https:\/\/media.kasperskydaily.com\/wp-content\/uploads\/sites\/91\/2024\/05\/31172427\/dense-pose-recognition-from-wi-fi-signal-03.jpeg\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-12381\" class=\"size-full wp-image-12381\" src=\"https:\/\/media.kasperskydaily.com\/wp-content\/uploads\/sites\/91\/2024\/05\/31172427\/dense-pose-recognition-from-wi-fi-signal-03.jpeg\" alt=\"DensePose: Foto\u011fraflarda poz alg\u0131lama \u00f6rnekleri \" width=\"2280\" height=\"1466\"><\/a><p id=\"caption-attachment-12381\" class=\"wp-caption-text\">DensePose modeli son derece s\u0131ra d\u0131\u015f\u0131 pozlar i\u00e7in bile iyi \u00e7al\u0131\u015f\u0131yor.<a href=\"https:\/\/arxiv.org\/pdf\/1802.00434.pdf\" target=\"_blank\" rel=\"nofollow noopener\"> Kaynak<\/a><\/p><\/div>\n<p>DensePose\u2019un bir di\u011fer avantaj\u0131 da \u00e7al\u0131\u015fmak i\u00e7in ola\u011fan\u00fcst\u00fc bir bilgi i\u015flem g\u00fcc\u00fc gerektirmemesidir. DensePose, GeForce GTX 1080 kullanarak (\u00e7al\u0131\u015fman\u0131n yay\u0131nland\u0131\u011f\u0131 tarihte bile pek \u00fcst d\u00fczey bir grafik kart\u0131 say\u0131lmazd\u0131) 240\u00d7320 \u00e7\u00f6z\u00fcn\u00fcrl\u00fckte saniyede 20-26 kare ve 800\u00d71100 \u00e7\u00f6z\u00fcn\u00fcrl\u00fckte saniyede be\u015f kareye kadar g\u00f6r\u00fcnt\u00fc yakalayabilir.<\/p>\n<h2>Wi-Fi \u00fczerinden DensePose: Foto\u011fraf yerine radyo dalgalar\u0131<\/h2>\n<p>Temel olarak Carnegie Mellon ara\u015ft\u0131rmac\u0131lar\u0131n\u0131n fikri, mevcut y\u00fcksek performansl\u0131 v\u00fccut alg\u0131lama yapay zeka modeli DensePose\u2019u kullanmak, ancak onu foto\u011fraf yerine Wi-Fi sinyalleriyle beslemekti.<\/p>\n<p>Deney i\u00e7in \u015f\u00f6yle bir d\u00fczenek kurdular:<\/p>\n<ul>\n<li>Her biri \u00fc\u00e7 antenle donat\u0131lm\u0131\u015f ve biri verici di\u011feri ise al\u0131c\u0131 olarak g\u00f6rev yapan standart TP-Link ev y\u00f6nlendiricilerine sahip iki stant.<\/li>\n<li>Bu stantlar aras\u0131nda konumland\u0131r\u0131lm\u0131\u015f bir alg\u0131lama sahnesi.<\/li>\n<li>Al\u0131c\u0131 y\u00f6nlendiricinin yan\u0131ndaki bir standa monte edilen ve ara\u015ft\u0131rmac\u0131lar\u0131n Wi-Fi sinyallerini kullanarak alg\u0131lamay\u0131 ama\u00e7lad\u0131klar\u0131 ayn\u0131 sahneyi yakalayacak olan bir kamera.<\/li>\n<\/ul>\n<div id=\"attachment_12382\" style=\"width: 1914px\" class=\"wp-caption aligncenter\"><a href=\"https:\/\/media.kasperskydaily.com\/wp-content\/uploads\/sites\/91\/2024\/05\/31172521\/dense-pose-recognition-from-wi-fi-signal-04.jpeg\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-12382\" class=\"size-full wp-image-12382\" src=\"https:\/\/media.kasperskydaily.com\/wp-content\/uploads\/sites\/91\/2024\/05\/31172521\/dense-pose-recognition-from-wi-fi-signal-04.jpeg\" alt=\"Wi-Fi \u00fczerinden DensePose: Y\u00f6ntemin genel ilkeleri\" width=\"1904\" height=\"1178\"><\/a><p id=\"caption-attachment-12382\" class=\"wp-caption-text\">Wi-Fi kullanarak insan pozlar\u0131n\u0131 alg\u0131lamaya y\u00f6nelik test tezgah\u0131n\u0131n genel diyagram\u0131.<a href=\"https:\/\/arxiv.org\/pdf\/2301.00250.pdf\" target=\"_blank\" rel=\"nofollow noopener\"> Kaynak<\/a><\/p><\/div>\n<p>Ard\u0131ndan, al\u0131c\u0131 y\u00f6nlendiricinin yan\u0131na yerle\u015ftirilen kameray\u0131 kullanarak v\u00fccut konumlar\u0131n\u0131 belirleyen DensePose\u2019u \u00e7al\u0131\u015ft\u0131rd\u0131lar ve al\u0131c\u0131 y\u00f6nlendiriciden gelen Wi-Fi sinyaliyle \u00e7al\u0131\u015fan ba\u015fka bir sinir a\u011f\u0131n\u0131 e\u011fitmekle g\u00f6revlendirdiler. Bu sinyal \u00f6nceden i\u015flenmi\u015f ve daha g\u00fcvenilir alg\u0131lama i\u00e7in de\u011fi\u015ftirilmi\u015ftir ancak bunlar k\u00fc\u00e7\u00fck ayr\u0131nt\u0131lard\u0131r. \u00d6nemli olan nokta, ara\u015ft\u0131rmac\u0131lar\u0131n ger\u00e7ekten de Wi-Fi sinyallerini kullanarak insan bedenlerinin uzamsal konumlar\u0131n\u0131 do\u011fru bir \u015fekilde yeniden yap\u0131land\u0131ran yeni bir Wi-Fi-DensePose modeli olu\u015fturabilmi\u015f olmalar\u0131d\u0131r.<\/p>\n<div id=\"attachment_12383\" style=\"width: 1170px\" class=\"wp-caption aligncenter\"><a href=\"https:\/\/media.kasperskydaily.com\/wp-content\/uploads\/sites\/91\/2024\/05\/31172617\/dense-pose-recognition-from-wi-fi-signal-05.jpeg\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-12383\" class=\"size-full wp-image-12383\" src=\"https:\/\/media.kasperskydaily.com\/wp-content\/uploads\/sites\/91\/2024\/05\/31172617\/dense-pose-recognition-from-wi-fi-signal-05.jpeg\" alt=\"DensePose Wi-Fi \u00fczerinden sahneleri ba\u015far\u0131yla alg\u0131lad\u0131 \" width=\"1160\" height=\"1634\"><\/a><p id=\"caption-attachment-12383\" class=\"wp-caption-text\">Model, iyi ko\u015fullarda insan pozlar\u0131n\u0131 \u00e7ok iyi alg\u0131layabiliyor.<a href=\"https:\/\/arxiv.org\/pdf\/2301.00250.pdf\" target=\"_blank\" rel=\"nofollow noopener\"> Kaynak<\/a><\/p><\/div>\n<h2>Y\u00f6ntemin s\u0131n\u0131rlamalar\u0131<\/h2>\n<p>Ancak, hen\u00fcz \u201cBilim \u0130nsanlar\u0131 Wi-Fi Kullanarak Duvarlar\u0131n \u0130\u00e7ini G\u00f6rmeyi \u00d6\u011frendi\u201d gibi ba\u015fl\u0131klar atmak i\u00e7in acele etmeyelim. Her \u015feyden \u00f6nce, buradaki \u201cg\u00f6rme\u201d olduk\u00e7a soyuttur. Model asl\u0131nda insan v\u00fccudunu \u201cg\u00f6rmez\u201d, ancak dolayl\u0131 verilere dayanarak belirli bir olas\u0131l\u0131kla yerini ve duru\u015f pozisyonunu tahmin edebilir.<\/p>\n<p>Wi-Fi sinyallerini kullanarak herhangi bir \u015feyi en ince ayr\u0131nt\u0131s\u0131na kadar g\u00f6rselle\u015ftirmek karma\u015f\u0131k bir i\u015ftir. Bu durum, ara\u015ft\u0131rmac\u0131lar\u0131n insan bedeninden \u00e7ok daha basit nesnelerle deneyler yapt\u0131\u011f\u0131 <a href=\"https:\/\/web.ece.ucsb.edu\/~ymostofi\/papers\/PallaproluKoranyMostofi_RadarConf2023.pdf\" target=\"_blank\" rel=\"nofollow noopener\">benzer bir ba\u015fka \u00e7al\u0131\u015fmada da<\/a> ortaya konmu\u015ftur ve sonu\u00e7lar, en hafif tabirle, ideal olmaktan uzakt\u0131r.<\/p>\n<div id=\"attachment_12384\" style=\"width: 1950px\" class=\"wp-caption aligncenter\"><a href=\"https:\/\/media.kasperskydaily.com\/wp-content\/uploads\/sites\/91\/2024\/05\/31172721\/dense-pose-recognition-from-wi-fi-signal-06.jpeg\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-12384\" class=\"size-full wp-image-12384\" src=\"https:\/\/media.kasperskydaily.com\/wp-content\/uploads\/sites\/91\/2024\/05\/31172721\/dense-pose-recognition-from-wi-fi-signal-06.jpeg\" alt=\"Wi-Fi sinyali kullanarak nesneleri g\u00f6rselle\u015ftirme \" width=\"1940\" height=\"882\"><\/a><p id=\"caption-attachment-12384\" class=\"wp-caption-text\">Wi-Fi sinyali kullanarak nesneleri g\u00f6rselle\u015ftirme: kenarlar ne kadar az belirgin olursa, o kadar k\u00f6t\u00fc olur.<a href=\"https:\/\/web.ece.ucsb.edu\/~ymostofi\/papers\/PallaproluKoranyMostofi_RadarConf2023.pdf\" target=\"_blank\" rel=\"nofollow noopener\"> Kaynak<\/a><\/p><\/div>\n<p>Carnegie Mellon \u00dcniversitesi ara\u015ft\u0131rmac\u0131lar\u0131 taraf\u0131ndan olu\u015fturulan modelin, foto\u011fraflardaki pozlar\u0131 alg\u0131lamaya y\u00f6nelik orijinal y\u00f6ntemden \u00f6nemli \u00f6l\u00e7\u00fcde daha az do\u011fru sonu\u00e7lar verdi\u011fini ve ayr\u0131ca olduk\u00e7a ciddi \u201chal\u00fcsinasyonlar\u201d sergiledi\u011fini de belirtmek gerekir. Model, al\u0131\u015f\u0131lmad\u0131k pozlarda veya ikiden fazla ki\u015finin yer ald\u0131\u011f\u0131 sahnelerde \u00f6zellikle zorlanmaktad\u0131r.<\/p>\n<div id=\"attachment_12385\" style=\"width: 2600px\" class=\"wp-caption aligncenter\"><a href=\"https:\/\/media.kasperskydaily.com\/wp-content\/uploads\/sites\/91\/2024\/05\/31172811\/dense-pose-recognition-from-wi-fi-signal-07.jpeg\"><img loading=\"lazy\" decoding=\"async\" aria-describedby=\"caption-attachment-12385\" class=\"size-full wp-image-12385\" src=\"https:\/\/media.kasperskydaily.com\/wp-content\/uploads\/sites\/91\/2024\/05\/31172811\/dense-pose-recognition-from-wi-fi-signal-07.jpeg\" alt=\"Wi-Fi \u00fczerinden DensePose alg\u0131lama hatalar\u0131 \" width=\"2590\" height=\"732\"><\/a><p id=\"caption-attachment-12385\" class=\"wp-caption-text\">Wi-Fi-DensePose modeli, standart olmayan pozlar\u0131 veya \u00e7ok say\u0131da insan v\u00fccudunu tek bir sahnede ele alma konusunda iyi bir i\u015f \u00e7\u0131karm\u0131yor.<a href=\"https:\/\/arxiv.org\/pdf\/2301.00250.pdf\" target=\"_blank\" rel=\"nofollow noopener\"> Kaynak<\/a><\/p><\/div>\n<p>Buna ek olarak, \u00e7al\u0131\u015fmadaki test ko\u015fullar\u0131; basit, iyi alg\u0131lanm\u0131\u015f bir geometri, verici ve al\u0131c\u0131 aras\u0131nda net bir g\u00f6r\u00fc\u015f hatt\u0131, minimum radyo sinyali paraziti vb. titizlikle kontrol edildi. K\u0131sacas\u0131 ara\u015ft\u0131rmac\u0131lar her \u015feyi sahneye radyo dalgalar\u0131yla kolayca \u201cn\u00fcfuz edebilecekleri\u201d \u015fekilde ayarlad\u0131lar. Bu ideal senaryonun ger\u00e7ek d\u00fcnyada tekrarlanmas\u0131 pek m\u00fcmk\u00fcn de\u011fildir.<\/p>\n<p>Yani birilerinin Wi-Fi y\u00f6nlendiricinize girip evde yapt\u0131klar\u0131n\u0131z\u0131 izlemesi konusunda endi\u015feleniyorsan\u0131z, rahat olun. Evinizde endi\u015felenmeniz gereken bir \u015fey varsa, o da ev aletleridir. \u00d6rne\u011fin, <a href=\"https:\/\/www.kaspersky.com.tr\/blog\/pet-feeders-vulnerabilities\/11557\/\" target=\"_blank\" rel=\"noopener\">ak\u0131ll\u0131 evcil hayvan besleyicileri<\/a> ve hatta <a href=\"https:\/\/www.kaspersky.com.tr\/blog\/robot-toy-security-issue\/12077\/\" target=\"_blank\" rel=\"noopener\">\u00e7ocuk oyuncaklar\u0131<\/a> kameralara, mikrofonlara ve bulut ba\u011flant\u0131s\u0131na sahipken, <a href=\"https:\/\/www.kaspersky.com.tr\/blog\/robot-vacuum-privacy\/11292\/\" target=\"_blank\" rel=\"noopener\">robot elektrikli s\u00fcp\u00fcrgeler<\/a> karanl\u0131kta kusursuz \u00e7al\u0131\u015fan lidarlara ve hareket etme yetene\u011fine bile sahiptir.<\/p>\n<p>Ve hemen d\u0131\u015far\u0131da, ba\u015fka bir casus (hem de d\u00f6rt tekerlekli!) sizi bekliyor. <a href=\"https:\/\/www.kaspersky.com.tr\/blog\/spies-on-wheels-how-carmakers-sell-your-intimate-data\/11804\/\" target=\"_blank\" rel=\"noopener\">Toplad\u0131klar\u0131 bilgi miktar\u0131 a\u00e7\u0131s\u0131ndan<\/a> g\u00fcn\u00fcm\u00fcz otomobilleri ak\u0131ll\u0131 saatlerden, ak\u0131ll\u0131 hoparl\u00f6rlerden ve di\u011fer g\u00fcnl\u00fck cihazlardan kilometrelerce ileridedir.<\/p>\n<input type=\"hidden\" class=\"category_for_banner\" value=\"premium-geek\">\n","protected":false},"excerpt":{"rendered":"<p>Ara\u015ft\u0131rmac\u0131lar, Wi-Fi sinyallerini kullanarak kapal\u0131 mekanda bulunan insanlar\u0131n konumlar\u0131n\u0131 ve duru\u015f pozisyonlar\u0131n\u0131 alg\u0131lamay\u0131 ba\u015fard\u0131lar ve bunu yapmak i\u00e7in sadece s\u0131radan ev tipi y\u00f6nlendiricilerini ve makine \u00f6\u011frenimini kulland\u0131lar.<\/p>\n","protected":false},"author":2726,"featured_media":12376,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[995],"tags":[1425,2173,2734,1610,2623,990,174,1424],"class_list":{"0":"post-12375","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-technology","8":"tag-ai","9":"tag-goruntuler","10":"tag-kablosuz-aglar","11":"tag-makine-ogrenimi","12":"tag-sinir-aglari","13":"tag-teknolojiler","14":"tag-wi-fi","15":"tag-yapay-zeka"},"hreflang":[{"hreflang":"tr","url":"https:\/\/www.kaspersky.com.tr\/blog\/dense-pose-recognition-from-wi-fi-signal\/12375\/"},{"hreflang":"en-in","url":"https:\/\/www.kaspersky.co.in\/blog\/dense-pose-recognition-from-wi-fi-signal\/27427\/"},{"hreflang":"en-ae","url":"https:\/\/me-en.kaspersky.com\/blog\/dense-pose-recognition-from-wi-fi-signal\/22750\/"},{"hreflang":"ar","url":"https:\/\/me.kaspersky.com\/blog\/dense-pose-recognition-from-wi-fi-signal\/11677\/"},{"hreflang":"en-us","url":"https:\/\/usa.kaspersky.com\/blog\/dense-pose-recognition-from-wi-fi-signal\/30111\/"},{"hreflang":"en-gb","url":"https:\/\/www.kaspersky.co.uk\/blog\/dense-pose-recognition-from-wi-fi-signal\/27581\/"},{"hreflang":"es-mx","url":"https:\/\/latam.kaspersky.com\/blog\/dense-pose-recognition-from-wi-fi-signal\/27374\/"},{"hreflang":"es","url":"https:\/\/www.kaspersky.es\/blog\/dense-pose-recognition-from-wi-fi-signal\/30026\/"},{"hreflang":"it","url":"https:\/\/www.kaspersky.it\/blog\/dense-pose-recognition-from-wi-fi-signal\/28837\/"},{"hreflang":"ru","url":"https:\/\/www.kaspersky.ru\/blog\/dense-pose-recognition-from-wi-fi-signal\/37400\/"},{"hreflang":"x-default","url":"https:\/\/www.kaspersky.com\/blog\/dense-pose-recognition-from-wi-fi-signal\/51216\/"},{"hreflang":"fr","url":"https:\/\/www.kaspersky.fr\/blog\/dense-pose-recognition-from-wi-fi-signal\/21879\/"},{"hreflang":"pt-br","url":"https:\/\/www.kaspersky.com.br\/blog\/dense-pose-recognition-from-wi-fi-signal\/22615\/"},{"hreflang":"de","url":"https:\/\/www.kaspersky.de\/blog\/dense-pose-recognition-from-wi-fi-signal\/31268\/"},{"hreflang":"ja","url":"https:\/\/blog.kaspersky.co.jp\/dense-pose-recognition-from-wi-fi-signal\/36386\/"},{"hreflang":"ru-kz","url":"https:\/\/blog.kaspersky.kz\/dense-pose-recognition-from-wi-fi-signal\/27729\/"},{"hreflang":"en-au","url":"https:\/\/www.kaspersky.com.au\/blog\/dense-pose-recognition-from-wi-fi-signal\/33580\/"},{"hreflang":"en-za","url":"https:\/\/www.kaspersky.co.za\/blog\/dense-pose-recognition-from-wi-fi-signal\/33242\/"}],"acf":[],"banners":"","maintag":{"url":"https:\/\/www.kaspersky.com.tr\/blog\/tag\/makine-ogrenimi\/","name":"makine \u00f6\u011frenimi"},"_links":{"self":[{"href":"https:\/\/www.kaspersky.com.tr\/blog\/wp-json\/wp\/v2\/posts\/12375","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.kaspersky.com.tr\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.kaspersky.com.tr\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.kaspersky.com.tr\/blog\/wp-json\/wp\/v2\/users\/2726"}],"replies":[{"embeddable":true,"href":"https:\/\/www.kaspersky.com.tr\/blog\/wp-json\/wp\/v2\/comments?post=12375"}],"version-history":[{"count":4,"href":"https:\/\/www.kaspersky.com.tr\/blog\/wp-json\/wp\/v2\/posts\/12375\/revisions"}],"predecessor-version":[{"id":12379,"href":"https:\/\/www.kaspersky.com.tr\/blog\/wp-json\/wp\/v2\/posts\/12375\/revisions\/12379"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.kaspersky.com.tr\/blog\/wp-json\/wp\/v2\/media\/12376"}],"wp:attachment":[{"href":"https:\/\/www.kaspersky.com.tr\/blog\/wp-json\/wp\/v2\/media?parent=12375"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.kaspersky.com.tr\/blog\/wp-json\/wp\/v2\/categories?post=12375"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.kaspersky.com.tr\/blog\/wp-json\/wp\/v2\/tags?post=12375"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}