\u0130\u015faret diliyle ileti\u015fimi kolayla\u015ft\u0131rmak ve engelleri ortadan kald\u0131rmak m\u0131 istiyorsunuz? Yapay zeka ve Google’\u0131n Teachable Machine arac\u0131yla, Brezilya \u0130\u015faret Dili (Libras) hareketlerini tan\u0131yabilen bir sistem olu\u015fturmak art\u0131k \u00e7ok daha eri\u015filebilir ve pratik. Bu makale, s\u0131f\u0131rdan ba\u015flayarak kendi Libras tan\u0131y\u0131c\u0131n\u0131z\u0131 nas\u0131l geli\u015ftirece\u011finizi ad\u0131m ad\u0131m a\u00e7\u0131kl\u0131yor.<\/p>\n
G\u00fcn\u00fcm\u00fczde ileti\u015fim, insan hayat\u0131n\u0131n vazge\u00e7ilmez bir par\u00e7as\u0131d\u0131r. Ancak i\u015fitme engelli bireyler i\u00e7in bu durum, bazen ciddi zorluklar do\u011furabilir. Brezilya \u0130\u015faret Dili (Libras), Brezilya’daki i\u015fitme engelli toplulu\u011fun ana ileti\u015fim arac\u0131d\u0131r. Bu \u00f6zel dilin tan\u0131nmas\u0131 ve anla\u015f\u0131lmas\u0131, hem i\u015fitme engelli bireylerin toplumsal entegrasyonunu sa\u011flamak hem de genel olarak ileti\u015fimi zenginle\u015ftirmek a\u00e7\u0131s\u0131ndan hayati bir role sahiptir. Peki, yapay zeka destekli bir Libras tan\u0131y\u0131c\u0131 sistemi neden bu kadar kritik? Cevap, eri\u015filebilirlik, e\u011fitim ve sosyal kapsay\u0131c\u0131l\u0131kt\u0131r.<\/p>\n
Bir Libras tan\u0131y\u0131c\u0131, i\u015fitme engelli bireyler ile i\u015fiten bireyler aras\u0131ndaki ileti\u015fim k\u00f6pr\u00fclerini g\u00fc\u00e7lendirme potansiyeline sahiptir. \u00d6rne\u011fin, bir sa\u011fl\u0131k kurulu\u015funda, bankada, kamu hizmetlerinde veya e\u011fitim ortamlar\u0131nda, i\u015fitme engelli bir ki\u015finin i\u015faret dilini anlamayan bir g\u00f6revliyle kar\u015f\u0131la\u015fmas\u0131, \u00f6nemli bilgilere eri\u015fimini veya temel haklar\u0131n\u0131 kullanmas\u0131n\u0131 engelleyebilir. Bu gibi durumlarda, an\u0131nda \u00e7eviri yapabilen veya i\u015faretleri tan\u0131yabilen bir sistem, bu engelleri ortadan kald\u0131rarak daha kapsay\u0131c\u0131 bir toplum yap\u0131s\u0131 olu\u015fturmaya yard\u0131mc\u0131 olabilir.<\/p>\n
E\u011fitim alan\u0131nda, Libras tan\u0131y\u0131c\u0131lar interaktif \u00f6\u011frenme deneyimleri sunabilir. \u0130\u015fitme engelli \u00f6\u011frenciler, ders materyallerini daha iyi anlayabilmek i\u00e7in i\u015faret dilini kullanan bir sistemle etkile\u015fim kurabilirler. Ayr\u0131ca, i\u015faret dilini \u00f6\u011frenmek isteyen i\u015fiten bireyler i\u00e7in de pratik bir ara\u00e7 haline gelebilir. Bir kamera kar\u015f\u0131s\u0131nda i\u015faretler yaparak an\u0131nda geri bildirim almak, \u00f6\u011frenme s\u00fcrecini h\u0131zland\u0131r\u0131r ve daha e\u011flenceli hale getirir. Bu, geleneksel \u00f6\u011frenme y\u00f6ntemlerinin \u00f6tesine ge\u00e7erek ki\u015fiselle\u015ftirilmi\u015f bir e\u011fitim yakla\u015f\u0131m\u0131 sunar.<\/p>\n
Ayr\u0131ca, bu t\u00fcr teknolojiler acil durum senaryolar\u0131nda hayat kurtar\u0131c\u0131 olabilir. Bir kaza an\u0131nda veya h\u0131zl\u0131 ileti\u015fimin gerekti\u011fi durumlarda, i\u015fitme engelli bir bireyin h\u0131zl\u0131ca bilgi aktarmas\u0131 veya yard\u0131m \u00e7a\u011f\u0131rmas\u0131 zor olabilir. \u0130\u015fte burada, i\u015faret dilini anlayan bir mobil uygulama veya cihaz devreye girerek, bu hayati ileti\u015fimi m\u00fcmk\u00fcn k\u0131lar. B\u00f6ylece, acil servislerin i\u015fitme engelli bireylerle daha etkin bir \u015fekilde ileti\u015fim kurmas\u0131 sa\u011flan\u0131r.<\/p>\n
\u00d6zetle, Libras tan\u0131y\u0131c\u0131lar sadece teknolojik bir yenilik olman\u0131n \u00f6tesinde, toplumsal e\u015fitlik ve eri\u015filebilirlik ad\u0131na at\u0131lm\u0131\u015f \u00f6nemli bir ad\u0131md\u0131r. \u0130\u015fitme engelli bireylerin ba\u011f\u0131ms\u0131zl\u0131\u011f\u0131n\u0131 ve topluma aktif kat\u0131l\u0131m\u0131n\u0131 destekleyen bu sistemler, gelecekte daha kapsay\u0131c\u0131 ve anlay\u0131\u015fl\u0131 bir d\u00fcnya in\u015fa etmemize yard\u0131mc\u0131 olacakt\u0131r. Bu nedenle, Teachable Machine gibi ara\u00e7larla bu t\u00fcr sistemleri geli\u015ftirmek, b\u00fcy\u00fck bir potansiyel sunmaktad\u0131r.<\/p>\n
Libras tan\u0131y\u0131c\u0131 gibi geli\u015fmi\u015f bir sistem olu\u015fturmak i\u00e7in \u00f6ncelikle arkas\u0131ndaki temel teknolojileri anlamak \u00f6nemlidir. Yapay Zeka (YZ) ve Makine \u00d6\u011frenimi (M\u00d6), bu t\u00fcr sistemlerin kalbinde yer al\u0131r. Ayr\u0131ca, bu teknolojileri herkesin eri\u015febilece\u011fi hale getiren Teachable Machine gibi ara\u00e7lar da s\u00fcrecin kritik bir par\u00e7as\u0131d\u0131r.<\/p>\n
Yapay zeka, makinelerin insan benzeri zekay\u0131 taklit etme yetene\u011fidir; yani \u00f6\u011frenme, problem \u00e7\u00f6zme, karar verme ve alg\u0131lama gibi yetenekleri sergilemesidir. Makine \u00f6\u011frenimi ise YZ’nin bir alt k\u00fcmesidir ve makinelerin a\u00e7\u0131k\u00e7a programlanmadan verilerden \u00f6\u011frenmesini sa\u011flayan bir dizi algoritma ve istatistiksel model i\u00e7erir. G\u00f6r\u00fcnt\u00fc tan\u0131ma gibi g\u00f6revlerde, M\u00d6 modelleri devasa veri k\u00fcmelerindeki desenleri \u00f6\u011frenerek belirli nesneleri veya hareketleri tan\u0131mlayabilir. Bizim Libras tan\u0131y\u0131c\u0131 projemizde, modelimiz belirli i\u015faret dili hareketlerinin g\u00f6rsel \u00f6zelliklerini \u00f6\u011frenecek ve bunlar\u0131 do\u011fru Libras kelimeleriyle ili\u015fkilendirecektir.<\/p>\n
G\u00f6r\u00fcnt\u00fc tabanl\u0131 tan\u0131ma sistemlerinde genellikle derin \u00f6\u011frenme modelleri kullan\u0131l\u0131r. \u00d6zellikle Evri\u015fimsel Sinir A\u011flar\u0131 (CNN’ler), g\u00f6r\u00fcnt\u00fcdeki desenleri ve \u00f6zellikleri tan\u0131mada son derece etkilidir. Bu a\u011flar, g\u00f6r\u00fcnt\u00fcn\u00fcn farkl\u0131 katmanlar\u0131n\u0131 analiz ederek kenarlar\u0131, \u015fekilleri ve daha karma\u015f\u0131k yap\u0131lar\u0131 tan\u0131mlar. Bu sayede, el hareketlerinin incelikli farkl\u0131l\u0131klar\u0131n\u0131 ay\u0131rt edebilirler. Ancak, derin \u00f6\u011frenme modellerini s\u0131f\u0131rdan in\u015fa etmek ve e\u011fitmek genellikle y\u00fcksek hesaplama g\u00fcc\u00fc, derin teknik bilgi ve b\u00fcy\u00fck veri setleri gerektirir. \u0130\u015fte tam da bu noktada Teachable Machine devreye girer.<\/p>\n
Google taraf\u0131ndan geli\u015ftirilen Teachable Machine, makine \u00f6\u011frenimi modelleri olu\u015fturmay\u0131 herkes i\u00e7in kolayla\u015ft\u0131ran web tabanl\u0131 bir ara\u00e7t\u0131r. Kod yazma bilgisi gerektirmeden, kullan\u0131c\u0131lar\u0131n kendi g\u00f6r\u00fcnt\u00fc, ses veya poz modellerini h\u0131zl\u0131ca e\u011fitmesine olanak tan\u0131r. Amac\u0131, yapay zekay\u0131 demokratikle\u015ftirmek ve yarat\u0131c\u0131 projelerde kullan\u0131m\u0131n\u0131 te\u015fvik etmektir. Temelinde, \u00f6nceden e\u011fitilmi\u015f (pre-trained) bir modelin (genellikle transfer \u00f6\u011frenimi prensibiyle \u00e7al\u0131\u015fan) ince ayar\u0131n\u0131 yaparak yeni verilerle \u00f6zelle\u015ftirmenizi sa\u011flar.<\/p>\n
Teachable Machine’in \u00e7al\u0131\u015fma prensibi olduk\u00e7a basittir: Sizden veri toplaman\u0131z\u0131 (\u00f6rne\u011fin kamera arac\u0131l\u0131\u011f\u0131yla i\u015faret dili hareketlerinin videolar\u0131n\u0131 \u00e7ekmenizi) ve bu verileri “s\u0131n\u0131flara” (\u00f6rne\u011fin “Merhaba”, “Te\u015fekk\u00fcrler”, “Evet”) ay\u0131rman\u0131z\u0131 ister. Her s\u0131n\u0131f i\u00e7in yeterli say\u0131da \u00f6rnek toplad\u0131ktan sonra, aray\u00fczdeki “Train Model” d\u00fc\u011fmesine basars\u0131n\u0131z. Teachable Machine, sa\u011flad\u0131\u011f\u0131n\u0131z verilerle \u00f6nceden e\u011fitilmi\u015f temel modeli otomatik olarak g\u00fcnceller ve i\u015faretlerinizi tan\u0131mak i\u00e7in optimize eder. E\u011fitim s\u00fcreci tamamland\u0131\u011f\u0131nda, modelinizi do\u011frudan taray\u0131c\u0131da test edebilir ve ard\u0131ndan \u00e7e\u015fitli platformlara entegre etmek \u00fczere d\u0131\u015fa aktarabilirsiniz. Bu sayede, karma\u015f\u0131k kodlama s\u00fcre\u00e7lerine girmeden, g\u00fc\u00e7l\u00fc bir yapay zeka modeline sahip olursunuz.<\/p>\n
\u00d6zetle, Teachable Machine, Libras tan\u0131y\u0131c\u0131 gibi g\u00f6rsel tan\u0131ma projeleri i\u00e7in m\u00fckemmel bir ba\u015flang\u0131\u00e7 noktas\u0131d\u0131r. Teknik karma\u015f\u0131kl\u0131\u011f\u0131 ortadan kald\u0131rarak projenin fikir ve uygulama a\u015famas\u0131na odaklanman\u0131z\u0131 sa\u011flar. Bu sayede, daha fazla ki\u015fi yapay zekan\u0131n g\u00fcc\u00fcnden faydalanarak eri\u015filebilir ve yenilik\u00e7i \u00e7\u00f6z\u00fcmler \u00fcretebilir.<\/p>\n
Kendi Libras tan\u0131y\u0131c\u0131n\u0131z\u0131 olu\u015fturmak, Teachable Machine sayesinde d\u00fc\u015f\u00fcnd\u00fc\u011f\u00fcn\u00fczden \u00e7ok daha kolayd\u0131r. Bu b\u00f6l\u00fcmde, s\u00fcreci ad\u0131m ad\u0131m inceleyecek, veri toplama, model e\u011fitimi ve entegrasyon a\u015famalar\u0131n\u0131 detayland\u0131raca\u011f\u0131z. Bu rehber sayesinde, teknik bilginiz az olsa bile kendi yapay zeka modelinizi geli\u015ftirebileceksiniz.<\/p>\n
Teachable Machine’i a\u00e7t\u0131\u011f\u0131n\u0131zda, size \u00fc\u00e7 ana proje t\u00fcr\u00fc sunulur: G\u00f6r\u00fcnt\u00fc Projesi, Ses Projesi ve Poz Projesi. Libras hareketleri el ve v\u00fccut pozisyonlar\u0131n\u0131 i\u00e7erdi\u011fi i\u00e7in “G\u00f6r\u00fcnt\u00fc Projesi” (Image Project) veya “Poz Projesi” (Pose Project) uygun olacakt\u0131r. Ancak, el i\u015faretlerinin detaylar\u0131n\u0131 yakalamak i\u00e7in genellikle “G\u00f6r\u00fcnt\u00fc Projesi” daha esnek bir ba\u015flang\u0131\u00e7 noktas\u0131 sunar, \u00e7\u00fcnk\u00fc poz projesi iskelet tespiti \u00fczerine odaklan\u0131r. Biz burada “G\u00f6r\u00fcnt\u00fc Projesi” \u00fczerinden ilerleyelim.<\/p>\n
Veri toplama, modelinizin ba\u015far\u0131s\u0131 i\u00e7in en kritik ad\u0131md\u0131r. Her bir Libras i\u015fareti i\u00e7in ayr\u0131 bir “s\u0131n\u0131f” (class) olu\u015fturman\u0131z gerekecek. \u00d6rne\u011fin, “Merhaba”, “Te\u015fekk\u00fcrler”, “Evet”, “Hay\u0131r” gibi i\u015faretler i\u00e7in ayr\u0131 s\u0131n\u0131flar tan\u0131mlay\u0131n. Her s\u0131n\u0131fa ait i\u015faretin farkl\u0131 a\u00e7\u0131lardan, farkl\u0131 ayd\u0131nlatma ko\u015fullar\u0131nda ve hatta farkl\u0131 ki\u015filerin yapaca\u011f\u0131 \u015fekilde bolca \u00f6rne\u011fini kaydetmeniz \u00f6nemlidir. Ne kadar \u00e7e\u015fitli ve bol veri toplarsan\u0131z, modeliniz o kadar sa\u011flam ve genellenebilir olur.<\/p>\n
Verilerinizi toplad\u0131ktan ve her s\u0131n\u0131f\u0131 etiketledikten sonra, s\u0131ra modeli e\u011fitmeye gelir. Teachable Machine, bu s\u00fcreci tek bir d\u00fc\u011fmeye basarak otomatikle\u015ftirir. “Train Model” d\u00fc\u011fmesine t\u0131klad\u0131\u011f\u0131n\u0131zda, taray\u0131c\u0131n\u0131zda veya Google’\u0131n sunucular\u0131nda (e\u011fer isterseniz) bir e\u011fitim s\u00fcreci ba\u015flar. Bu s\u00fcre\u00e7, toplad\u0131\u011f\u0131n\u0131z g\u00f6r\u00fcnt\u00fcleri analiz eder ve her bir Libras i\u015faretinin benzersiz \u00f6zelliklerini \u00f6\u011frenir.<\/p>\n
Modeliniz beklendi\u011fi gibi \u00e7al\u0131\u015f\u0131yorsa, onu projelerinize entegre etmek i\u00e7in d\u0131\u015fa aktarabilirsiniz. Teachable Machine, farkl\u0131 kullan\u0131m senaryolar\u0131 i\u00e7in \u00e7e\u015fitli d\u0131\u015fa aktarma se\u00e7enekleri sunar:<\/p>\n
\u00d6rne\u011fin, bir web uygulamas\u0131na entegrasyon i\u00e7in TensorFlow.js se\u00e7ene\u011fini kullanabilirsiniz. “Export Model” d\u00fc\u011fmesine t\u0131klad\u0131\u011f\u0131n\u0131zda, size \u00f6zel bir kod par\u00e7ac\u0131\u011f\u0131 ve model dosyalar\u0131 i\u00e7eren bir ba\u011flant\u0131 verilir. \u0130\u015fte bu kodu HTML dosyan\u0131za nas\u0131l ekleyece\u011finize dair basit bir \u00f6rnek:<\/p>\n
\n\n\n\n \n \n Libras Tan\u0131y\u0131c\u0131 Uygulamas\u0131<\/title>\n <script src=\"https:\/\/cdn.jsdelivr.net\/npm\/@tensorflow\/tfjs@latest\/dist\/tf.min.js\"><\/script>\n <script src=\"https:\/\/cdn.jsdelivr.net\/npm\/@teachablemachine\/image@latest\/dist\/teachablemachine-image.min.js\"><\/script>\n<\/head>\n<body>\n <h1>Libras \u0130\u015faret Dili Tan\u0131y\u0131c\u0131<\/h1>\n <video id=\"webcam\" autoplay playsinline><\/video>\n <div id=\"label-container\"><\/div>\n\n <script type=\"text\/javascript\">\n const URL = \"https:\/\/teachablemachine.withgoogle.com\/models\/YOUR_MODEL_ID\/\"; \/\/ Model ID'nizi buraya yap\u0131\u015ft\u0131r\u0131n\n let model, webcam, labelContainer, maxPredictions;\n\n async function init() {\n const modelURL = URL + \"model.json\";\n const metadataURL = URL + \"metadata.json\";\n\n model = await tmImage.load(modelURL, metadataURL);\n maxPredictions = model.getTotalClasses();\n\n const flip = true;\n webcam = new tmImage.Webcam(200, 200, flip); \/\/ Geni\u015flik, Y\u00fckseklik, Flip\n await webcam.setup();\n await webcam.play();\n window.requestAnimationFrame(loop);\n\n document.getElementById(\"webcam\").appendChild(webcam.canvas);\n labelContainer = document.getElementById(\"label-container\");\n for (let i = 0; i < maxPredictions; i++) {\n labelContainer.appendChild(document.createElement(\"div\"));\n }\n }\n\n async function loop() {\n webcam.update();\n await predict();\n window.requestAnimationFrame(loop);\n }\n\n async function predict() {\n const prediction = await model.predict(webcam.canvas);\n for (let i = 0; i < maxPredictions; i++) {\n const classPrediction =\n prediction[i].className + \": \" + prediction[i].probability.toFixed(2);\n labelContainer.childNodes[i].innerHTML = classPrediction;\n }\n }\n\n init(); \/\/ Uygulamay\u0131 ba\u015flat\n <\/script>\n<\/body>\n<html>\n<\/pre>\n<p><\/code><\/p>\n<p>Bu kod blo\u011fu, web kameran\u0131zdan gelen g\u00f6r\u00fcnt\u00fcleri al\u0131r, modelinize besler ve tan\u0131nan Libras i\u015faretini ekranda g\u00f6sterir. <code>YOUR_MODEL_ID<\/code> k\u0131sm\u0131n\u0131 Teachable Machine'den d\u0131\u015fa aktard\u0131\u011f\u0131n\u0131zda size verilen model kimli\u011fi ile de\u011fi\u015ftirmeniz gerekmektedir. Bu sayede, kendi interaktif Libras tan\u0131y\u0131c\u0131n\u0131z\u0131 h\u0131zl\u0131ca bir web sayfas\u0131 \u00fczerinde \u00e7al\u0131\u015f\u0131r hale getirebilirsiniz.<\/p>\n<h2>Model Performans\u0131n\u0131 Art\u0131rmak i\u00e7in \u0130pu\u00e7lar\u0131 ve P\u00fcf Noktalar\u0131 Nelerdir?<\/h2>\n<p>Bir Libras tan\u0131y\u0131c\u0131 modelini Teachable Machine ile olu\u015fturmak kolayd\u0131r, ancak modelin ger\u00e7ek d\u00fcnya senaryolar\u0131nda y\u00fcksek do\u011frulukla \u00e7al\u0131\u015fmas\u0131n\u0131 sa\u011flamak biraz daha \u00f6zen gerektirir. \u0130\u015fte model performans\u0131n\u0131 art\u0131rmak i\u00e7in uygulayabilece\u011finiz baz\u0131 ileri d\u00fczey ipu\u00e7lar\u0131 ve p\u00fcf noktalar\u0131:<\/p>\n<h3>1. Veri Seti \u00c7e\u015fitlili\u011fini ve Miktar\u0131n\u0131 Art\u0131rma<\/h3>\n<p>Modelinizin genelleme yetene\u011fi, e\u011fitim verilerinizin kalitesine ve \u00e7e\u015fitlili\u011fine do\u011frudan ba\u011fl\u0131d\u0131r. Sadece birka\u00e7 ki\u015fiyle, tek bir arka planda ve ayd\u0131nlatmada \u00e7ekilen verilerle e\u011fitilen bir model, yeni bir ortamda veya farkl\u0131 bir ki\u015fi taraf\u0131ndan yap\u0131lan i\u015faretleri tan\u0131makta zorlanabilir.<\/p>\n<ul>\n<li><strong>Farkl\u0131 Ki\u015filer:<\/strong> M\u00fcmk\u00fcnse, farkl\u0131 ya\u015f, cinsiyet ve ten rengine sahip ki\u015filerin i\u015faretleri yapmas\u0131n\u0131 sa\u011flay\u0131n. Her bireyin i\u015faret yap\u0131\u015f \u015fekli k\u00fc\u00e7\u00fck farkl\u0131l\u0131klar g\u00f6sterebilir.<\/li>\n<li><strong>De\u011fi\u015fen Arka Planlar:<\/strong> Modelin sadece i\u015farete odaklanmas\u0131n\u0131 sa\u011flamak i\u00e7in, farkl\u0131 arka planlarda (d\u00fcz renk duvarlar, da\u011f\u0131n\u0131k odalar, a\u00e7\u0131k hava vb.) \u00e7ekim yap\u0131n.<\/li>\n<li><strong>Farkl\u0131 Ayd\u0131nlatma Ko\u015fullar\u0131:<\/strong> D\u00fc\u015f\u00fck \u0131\u015f\u0131k, parlak \u0131\u015f\u0131k, g\u00f6lgeli alanlar gibi farkl\u0131 ayd\u0131nlatma ko\u015fullar\u0131nda \u00f6rnekler toplay\u0131n.<\/li>\n<li><strong>A\u00e7\u0131 \u00c7e\u015fitlili\u011fi:<\/strong> \u0130\u015faretleri farkl\u0131 a\u00e7\u0131lardan (hafif\u00e7e soldan, sa\u011fdan, yukar\u0131dan a\u015fa\u011f\u0131dan) \u00e7ekmek, modelin i\u015faretin \u00fc\u00e7 boyutlu yap\u0131s\u0131n\u0131 daha iyi anlamas\u0131na yard\u0131mc\u0131 olur.<\/li>\n<li><strong>Veri \u00c7o\u011faltma (Data Augmentation):<\/strong> Mevcut verilerinizi yapay olarak art\u0131rabilirsiniz. G\u00f6r\u00fcnt\u00fcleri hafif\u00e7e d\u00f6nd\u00fcrmek, yak\u0131nla\u015ft\u0131rmak\/uzakla\u015ft\u0131rmak, yatay \u00e7evirmek veya renk tonlar\u0131n\u0131 de\u011fi\u015ftirmek gibi i\u015flemlerle yeni e\u011fitim \u00f6rnekleri olu\u015fturulabilir. Teachable Machine do\u011frudan veri \u00e7o\u011faltma yapmasa da, foto\u011fraflar\u0131n\u0131z\u0131 \u00e7ekip d\u00fczenleyerek bu \u00e7e\u015fitlili\u011fi manuel olarak sa\u011flayabilirsiniz.<\/li>\n<\/ul>\n<p>Her s\u0131n\u0131f i\u00e7in minimum 200-300 hatta daha fazla \u00f6rnek hedeflemek, modelinizin daha robust (sa\u011flam) olmas\u0131n\u0131 sa\u011flayacakt\u0131r. Veri ne kadar zengin olursa, model o kadar az hata yapar.<\/p>\n<h3>2. E\u011fitim Ayarlar\u0131n\u0131 Optimize Etme<\/h3>\n<p>Teachable Machine'deki \"Advanced\" (Geli\u015fmi\u015f) ayarlar b\u00f6l\u00fcm\u00fc, modelinizin \u00f6\u011frenme s\u00fcrecini etkileyen baz\u0131 parametreleri de\u011fi\u015ftirmenize olanak tan\u0131r. Bu ayarlar, modelinizin performans\u0131n\u0131 \u00f6nemli \u00f6l\u00e7\u00fcde etkileyebilir:<\/p>\n<ul>\n<li><strong>Epochs:<\/strong> Modelin t\u00fcm e\u011fitim veri setini ka\u00e7 kez g\u00f6rece\u011fini belirtir. Daha y\u00fcksek bir epoch de\u011feri, modelin daha fazla \u00f6\u011frenme f\u0131rsat\u0131 bulmas\u0131n\u0131 sa\u011flar, ancak ayn\u0131 zamanda a\u015f\u0131r\u0131 \u00f6\u011frenmeye (overfitting) yol a\u00e7abilir. Genellikle 50-100 aras\u0131 bir de\u011ferle ba\u015flay\u0131p, ihtiya\u00e7 duyulursa art\u0131rabilirsiniz.<\/li>\n<li><strong>Batch Size:<\/strong> Her e\u011fitim ad\u0131m\u0131nda i\u015flenecek \u00f6rnek say\u0131s\u0131n\u0131 belirler. Daha k\u00fc\u00e7\u00fck batch size'lar daha kararl\u0131 bir \u00f6\u011frenme sa\u011flayabilirken, daha b\u00fcy\u00fck batch size'lar daha h\u0131zl\u0131 e\u011fitime yol a\u00e7abilir.<\/li>\n<li><strong>Learning Rate (\u00d6\u011frenme Oran\u0131):<\/strong> Modelin her e\u011fitim ad\u0131m\u0131nda ne kadar h\u0131zl\u0131 \u00f6\u011frendi\u011fini kontrol eder. \u00c7ok y\u00fcksek bir \u00f6\u011frenme oran\u0131, modelin optimal noktay\u0131 \"atlamas\u0131na\" neden olabilirken, \u00e7ok d\u00fc\u015f\u00fck bir oran e\u011fitimin \u00e7ok yava\u015f olmas\u0131na yol a\u00e7abilir. Genellikle <code>0.001<\/code> gibi k\u00fc\u00e7\u00fck de\u011ferlerle ba\u015flan\u0131r.<\/li>\n<\/ul>\n<p>Bu parametrelerle deneme yapmak, modelinizin belirli veri setiniz i\u00e7in en iyi performans\u0131 g\u00f6stermesini sa\u011flayacakt\u0131r. Her de\u011fi\u015fiklikten sonra modeli yeniden e\u011fitin ve performans\u0131n\u0131 test edin.<\/p>\n<h3>3. Negatif \u00d6rnekler ve Yanl\u0131\u015f Pozitifleri Azaltma<\/h3>\n<p>Modelin sadece \u00f6\u011frenmek istedi\u011finiz i\u015faretleri de\u011fil, ayn\u0131 zamanda i\u015faret yapmad\u0131\u011f\u0131n\u0131z durumlar\u0131 da \"bilmesi\" \u00f6nemlidir. Bu, yanl\u0131\u015f pozitifleri (yani bir i\u015faret yapmad\u0131\u011f\u0131n\u0131z halde modelin bir i\u015faret tan\u0131mas\u0131) azaltmaya yard\u0131mc\u0131 olur.<\/p>\n<ul>\n<li><strong>\"Arka Plan\" veya \"Hi\u00e7bir \u015eey\" S\u0131n\u0131f\u0131:<\/strong> Eliniz serbest dururken, bilgisayar\u0131n bo\u015f ekran\u0131n\u0131 veya genel arka plan\u0131 g\u00f6steren bir s\u0131n\u0131f olu\u015fturun. Bu s\u0131n\u0131fa ait bolca \u00f6rnek eklemek, modelin i\u015faret olmad\u0131\u011f\u0131nda \"hi\u00e7bir \u015fey\" olarak s\u0131n\u0131fland\u0131rmas\u0131n\u0131 sa\u011flayacakt\u0131r. Bu sayede, model gereksiz yere herhangi bir i\u015faret alg\u0131lamaktan ka\u00e7\u0131n\u0131r ve kullan\u0131c\u0131 deneyimini iyile\u015ftirir.<\/li>\n<\/ul>\n<h3>4. Erken Durdurma (Early Stopping)<\/h3>\n<p>Overfitting, modelin e\u011fitim verilerini ezberlemesi ancak yeni, bilinmeyen verilere genelleme yapamamas\u0131 durumudur. Teachable Machine do\u011frudan erken durdurma se\u00e7ene\u011fi sunmasa da, modelinizi canl\u0131 \u00f6nizlemede test ederek bu durumu g\u00f6zlemleyebilirsiniz. E\u011fer e\u011fitim do\u011frulu\u011fu y\u00fckselirken test do\u011frulu\u011fu d\u00fc\u015fmeye ba\u015fl\u0131yorsa, model overfitting olmaya ba\u015flam\u0131\u015f olabilir. Bu durumda, daha az epoch ile veya daha fazla veri \u00e7e\u015fitlili\u011fi ile tekrar denemek faydal\u0131 olacakt\u0131r.<\/p>\n<p>Bu ipu\u00e7lar\u0131, Teachable Machine ile olu\u015fturdu\u011funuz Libras tan\u0131y\u0131c\u0131 modelinizin do\u011frulu\u011funu ve g\u00fcvenilirli\u011fini \u00f6nemli \u00f6l\u00e7\u00fcde art\u0131rman\u0131za yard\u0131mc\u0131 olacakt\u0131r. Unutmay\u0131n, iyi bir yapay zeka modeli, iyi bir veri setine ve dikkatli bir deneme-yan\u0131lma s\u00fcrecine dayan\u0131r.<\/p>\n<h2>Mobil Uygulamalar i\u00e7in Entegrasyon ve Optimizasyon: Her Yerde Eri\u015filebilir Libras<\/h2>\n<p>Bir Libras tan\u0131y\u0131c\u0131 modelini olu\u015fturmak harika bir ba\u015flang\u0131\u00e7t\u0131r, ancak bu modelin ger\u00e7ek d\u00fcnyada kullan\u0131labilir olmas\u0131 i\u00e7in \u00e7e\u015fitli platformlara entegre edilmesi gerekir. \u00d6zellikle mobil uygulamalar, Libras tan\u0131y\u0131c\u0131lar\u0131n\u0131n eri\u015filebilirli\u011fini maksimum d\u00fczeye \u00e7\u0131karabilir. Cep telefonlar\u0131 ve tabletler \u00fczerindeki uygulamalar, kullan\u0131c\u0131lar\u0131n her an ve her yerde i\u015faret dili \u00e7evirisi yapmas\u0131na olanak tan\u0131r. Peki, Teachable Machine ile e\u011fitilen bir modeli mobil platformlara nas\u0131l entegre edebilir ve optimize edebiliriz?<\/p>\n<h3>Mobil Entegrasyon Se\u00e7enekleri<\/h3>\n<p>Teachable Machine, mobil entegrasyon i\u00e7in \u00e7e\u015fitli formatlar sunar:<\/p>\n<ol>\n<li><strong>TensorFlow Lite:<\/strong> Android ve iOS cihazlar i\u00e7in idealdir. TensorFlow Lite, modelleri mobil cihazlarda verimli bir \u015fekilde \u00e7al\u0131\u015facak \u015fekilde optimize eder. Daha k\u00fc\u00e7\u00fck boyutlu modeller, daha d\u00fc\u015f\u00fck gecikme s\u00fcreleri ve daha az g\u00fc\u00e7 t\u00fcketimi sa\u011flar. Teachable Machine'den modelinizi TensorFlow Lite format\u0131nda d\u0131\u015fa aktarabilir, ard\u0131ndan Android Studio (Kotlin\/Java) veya Xcode (Swift\/Objective-C) kullanarak uygulaman\u0131za entegre edebilirsiniz. Bu, cihaz \u00fczerinde (on-device) \u00e7\u0131kar\u0131m yaparak internet ba\u011flant\u0131s\u0131 olmadan bile \u00e7al\u0131\u015fabilen uygulamalar olu\u015fturman\u0131za olanak tan\u0131r.<\/li>\n<li><strong>Core ML (Yaln\u0131zca iOS):<\/strong> Apple ekosistemi i\u00e7in optimize edilmi\u015f bir formatt\u0131r. Core ML, iPhone ve iPad'lerde makine \u00f6\u011frenimi modellerinin h\u0131zl\u0131 ve verimli \u00e7al\u0131\u015fmas\u0131n\u0131 sa\u011flar. E\u011fer hedef kitleniz a\u011f\u0131rl\u0131kl\u0131 olarak iOS kullan\u0131c\u0131lar\u0131ysa, Core ML entegrasyonu performans\u0131 art\u0131rabilir.<\/li>\n<li><strong>Web Tabanl\u0131 \u00c7\u00f6z\u00fcmler (PWA'lar):<\/strong> TensorFlow.js ile e\u011fitilen modeli bir Web uygulamas\u0131na entegre etmek ve bu uygulamay\u0131 PWA (Progressive Web App) olarak yay\u0131nlamak, hem Android hem de iOS cihazlarda \u00e7al\u0131\u015fabilen \u00e7apraz platform bir \u00e7\u00f6z\u00fcm sunar. Bu yakla\u015f\u0131m, uygulama ma\u011fazalar\u0131na ba\u011f\u0131ml\u0131l\u0131\u011f\u0131 azalt\u0131r ve g\u00fcncellemeleri kolayla\u015ft\u0131r\u0131r. Taray\u0131c\u0131 tabanl\u0131 \u00e7al\u0131\u015ft\u0131\u011f\u0131 i\u00e7in, performans biraz daha d\u00fc\u015f\u00fck olabilir ancak entegrasyonu genellikle daha h\u0131zl\u0131d\u0131r.<\/li>\n<\/ol>\n<h3>Mobil Optimizasyon \u0130pu\u00e7lar\u0131<\/h3>\n<p>Mobil cihazlar\u0131n s\u0131n\u0131rl\u0131 kaynaklar\u0131 g\u00f6z \u00f6n\u00fcne al\u0131nd\u0131\u011f\u0131nda, modelinizi ve uygulaman\u0131z\u0131 optimize etmek kritik \u00f6neme sahiptir:<\/p>\n<ul>\n<li><strong>Model Boyutu:<\/strong> Teachable Machine genellikle optimize edilmi\u015f modeller \u00fcretse de, m\u00fcmk\u00fcn oldu\u011funca az s\u0131n\u0131f ve sadece gerekli verilerle e\u011fitilmi\u015f modeller kullanmak, model boyutunu d\u00fc\u015f\u00fcr\u00fcr. Daha k\u00fc\u00e7\u00fck modeller, daha h\u0131zl\u0131 y\u00fckleme ve daha az depolama alan\u0131 demektir.<\/li>\n<li><strong>Cihaz \u00dczerinde \u00c7\u0131kar\u0131m (On-Device Inference):<\/strong> \u0130nternet ba\u011flant\u0131s\u0131na ba\u011f\u0131ml\u0131l\u0131\u011f\u0131 ortadan kald\u0131rmak i\u00e7in modeli do\u011frudan cihaz \u00fczerinde \u00e7al\u0131\u015ft\u0131rmay\u0131 tercih edin. Bu, gizlili\u011fi art\u0131r\u0131r ve gecikme s\u00fcresini azalt\u0131r.<\/li>\n<li><strong>Kamera Ayarlar\u0131:<\/strong> Mobil uygulamada kamera \u00e7\u00f6z\u00fcn\u00fcrl\u00fc\u011f\u00fcn\u00fc ve kare h\u0131z\u0131n\u0131 (FPS) ayarlayarak performans dengesini sa\u011flay\u0131n. Y\u00fcksek \u00e7\u00f6z\u00fcn\u00fcrl\u00fckler daha iyi do\u011fruluk sunarken, performans\u0131 d\u00fc\u015f\u00fcrebilir.<\/li>\n<li><strong>Kullan\u0131c\u0131 Aray\u00fcz\u00fc (UI) ve Deneyimi (UX):<\/strong> Uygulaman\u0131n kullan\u0131m\u0131 kolay olmal\u0131, i\u015faret dili yapan ki\u015fiye net geri bildirimler sunmal\u0131 ve i\u015faretin tan\u0131nma olas\u0131l\u0131\u011f\u0131n\u0131 anla\u015f\u0131l\u0131r bir \u015fekilde g\u00f6stermelidir. H\u0131zl\u0131 geri bildirim, kullan\u0131c\u0131 memnuniyetini art\u0131r\u0131r.<\/li>\n<\/ul>\n<h3>Web Tabanl\u0131 Uygulamalar i\u00e7in Responsive Tasar\u0131m \u00d6rne\u011fi<\/h3>\n<p>E\u011fer TensorFlow.js ile bir web uygulamas\u0131 geli\u015ftiriyorsan\u0131z, uygulaman\u0131z\u0131n farkl\u0131 ekran boyutlar\u0131nda (mobil, tablet, masa\u00fcst\u00fc) iyi g\u00f6r\u00fcnmesini sa\u011flamak i\u00e7in duyarl\u0131 (responsive) tasar\u0131m kullanmal\u0131s\u0131n\u0131z. CSS medya sorgular\u0131 (media queries) bu konuda en g\u00fc\u00e7l\u00fc ara\u00e7lardan biridir:<\/p>\n<pre><code class=\"language-css\">\n\/* Genel stil ayarlar\u0131 *\/\nbody {\n font-family: Arial, sans-serif;\n display: flex;\n flex-direction: column;\n align-items: center;\n margin: 20px;\n background-color: #f4f4f4;\n color: #333;\n}\n\nh1 {\n color: #007bff;\n margin-bottom: 30px;\n}\n\nvideo {\n width: 100%; \/* Mobil cihazlarda tam geni\u015flik *\/\n max-width: 320px; \/* Belirli bir geni\u015fli\u011fi a\u015fmas\u0131n *\/\n height: auto;\n border: 2px solid #ddd;\n border-radius: 8px;\n margin-bottom: 20px;\n}\n\n#label-container {\n width: 100%;\n max-width: 320px;\n text-align: center;\n font-size: 1.2em;\n padding: 15px;\n border-radius: 8px;\n background-color: #fff;\n box-shadow: 0 2px 4px rgba(0,0,0,0.1);\n}\n\n\/* Tablet ve k\u00fc\u00e7\u00fck masa\u00fcst\u00fc ekranlar i\u00e7in *\/\n@media (min-width: 600px) {\n video {\n max-width: 480px;\n }\n #label-container {\n max-width: 480px;\n font-size: 1.3em;\n }\n}\n\n\/* B\u00fcy\u00fck masa\u00fcst\u00fc ekranlar i\u00e7in *\/\n@media (min-width: 992px) {\n video {\n max-width: 640px;\n }\n #label-container {\n max-width: 640px;\n font-size: 1.4em;\n }\n}\n<\/pre>\n<p><\/code><\/p>\n<p>Yukar\u0131daki CSS \u00f6rne\u011fi, <code><video><\/code> elementi ve <code>#label-container<\/code> elementinin farkl\u0131 ekran boyutlar\u0131nda nas\u0131l adapte olaca\u011f\u0131n\u0131 g\u00f6stermektedir. Mobil cihazlarda tam geni\u015flikte ba\u015flarken, daha b\u00fcy\u00fck ekranlarda belirli bir <code>max-width<\/code> de\u011ferine ula\u015farak i\u00e7eri\u011fin daha d\u00fczenli g\u00f6r\u00fcnmesini sa\u011flar. Bu, kullan\u0131c\u0131n\u0131n cihaz\u0131 ne olursa olsun, uygulaman\u0131z\u0131n tutarl\u0131 ve eri\u015filebilir bir deneyim sunmas\u0131na yard\u0131mc\u0131 olur.<\/p>\n<p>Mobil entegrasyon ve optimizasyon, Libras tan\u0131y\u0131c\u0131 projenizin ger\u00e7ek potansiyelini ortaya \u00e7\u0131kar\u0131r. Bu sayede, i\u015faret dili \u00e7evirisi teknolojisi, ihtiya\u00e7 duyan herkesin cebine girerek, ileti\u015fimi daha kapsay\u0131c\u0131 hale getirebilir.<\/p>\n<h2>Vaka Analizi: Ger\u00e7ek D\u00fcnya Uygulamalar\u0131 ve Ba\u015far\u0131 Hikayeleri<\/h2>\n<p>Teachable Machine ile e\u011fitilmi\u015f bir Libras tan\u0131y\u0131c\u0131n\u0131n potansiyeli sadece teknik bir ba\u015far\u0131dan ibaret de\u011fildir; ger\u00e7ek d\u00fcnyada somut faydalar sa\u011flayarak insanlar\u0131n hayatlar\u0131n\u0131 de\u011fi\u015ftirebilir. \u0130\u015fte bu teknolojinin uygulanabilece\u011fi baz\u0131 vaka analizleri ve ilham verici senaryolar:<\/p>\n<h3>1. E\u011fitimde \u0130nteraktif \u00d6\u011frenme Asistanlar\u0131<\/h3>\n<p>Bir okulda, i\u015fitme engelli \u00e7ocuklar\u0131n i\u015faret dili \u00f6\u011frenme s\u00fcre\u00e7lerini desteklemek i\u00e7in bir Libras tan\u0131y\u0131c\u0131 uygulamas\u0131 geli\u015ftirildi. Uygulama, \u00e7ocuklar\u0131n bir kamera kar\u015f\u0131s\u0131nda yapt\u0131klar\u0131 i\u015faretleri an\u0131nda tan\u0131yarak do\u011fru veya yanl\u0131\u015f geri bildirim veriyor. \u00d6rne\u011fin, \"ev\" i\u015faretini do\u011fru yapt\u0131klar\u0131nda, ekranda bir ev resmi beliriyor ve bir alk\u0131\u015f sesi (i\u015fitenler i\u00e7in) veya titre\u015fim (i\u015fitme engelliler i\u00e7in) geliyor. Yanl\u0131\u015f yapt\u0131klar\u0131nda ise, do\u011fru i\u015fareti g\u00f6steren bir video klip oynat\u0131l\u0131yor. Bu interaktif sistem, \u00e7ocuklar\u0131n kendi h\u0131zlar\u0131nda \u00f6\u011frenmelerini sa\u011fl\u0131yor ve \u00f6\u011fretmenlerin bireysel \u00f6\u011frenme ihtiya\u00e7lar\u0131na daha fazla odaklanmalar\u0131na olanak tan\u0131yor.<\/p>\n<p>Bu vaka analizinde, Teachable Machine sayesinde h\u0131zl\u0131ca prototip olu\u015fturuldu. \u00d6\u011fretmenler ve \u00f6zel e\u011fitim uzmanlar\u0131, \u00f6\u011frencilerin en s\u0131k zorland\u0131\u011f\u0131 i\u015faretleri belirleyerek veri toplama s\u00fcrecine do\u011frudan kat\u0131ld\u0131. Model, \u00e7ocuklar\u0131n tipik el hareketlerini ve hafif varyasyonlar\u0131n\u0131 \u00f6\u011frenmek \u00fczere e\u011fitildi. Sonu\u00e7 olarak, \u00f6\u011frencilerin i\u015faret dili becerileri daha h\u0131zl\u0131 geli\u015fti ve derse kat\u0131l\u0131mlar\u0131 \u00f6nemli \u00f6l\u00e7\u00fcde artt\u0131.<\/p>\n<h3>2. Kamu Hizmetlerinde Eri\u015filebilir \u0130leti\u015fim Noktalar\u0131<\/h3>\n<p>B\u00fcy\u00fck bir belediye binas\u0131n\u0131n dan\u0131\u015fma masas\u0131na, i\u015fitme engelli vatanda\u015flar\u0131n belediye personeliyle daha rahat ileti\u015fim kurabilmesi i\u00e7in bir Libras tan\u0131y\u0131c\u0131 kiosk entegre edildi. Kiosk, bir kamera ve dokunmatik bir ekrandan olu\u015fuyordu. Vatanda\u015flar kameraya Libras i\u015faretleri yaparak sorular\u0131n\u0131 sorabiliyor, sistem bu i\u015faretleri metne \u00e7evirip ekranda g\u00f6steriyordu. Personel de ayn\u0131 \u015fekilde yaz\u0131l\u0131 yan\u0131tlar vererek veya \u00f6nceden tan\u0131mlanm\u0131\u015f Libras i\u015faretlerini sistemin yapay sesle (text-to-speech) okumas\u0131n\u0131 sa\u011flayarak ileti\u015fim kurabiliyordu.<\/p>\n<p>Bu proje i\u00e7in, en yayg\u0131n kamu hizmeti terimlerini i\u00e7eren bir Libras veri seti olu\u015fturuldu. Teachable Machine, farkl\u0131 personellerin ve vatanda\u015flar\u0131n yapaca\u011f\u0131 i\u015faretlerin varyasyonlar\u0131n\u0131 \u00f6\u011frenmek i\u00e7in kullan\u0131ld\u0131. Bu \u00e7\u00f6z\u00fcm, i\u015fitme engelli vatanda\u015flar\u0131n kamu hizmetlerine eri\u015fimini kolayla\u015ft\u0131rman\u0131n yan\u0131 s\u0131ra, belediyenin toplumsal kapsay\u0131c\u0131l\u0131k konusundaki taahh\u00fcd\u00fcn\u00fc de g\u00fc\u00e7lendirdi. Sistem, \u00f6zellikle yo\u011fun saatlerde veya terc\u00fcman bulunmad\u0131\u011f\u0131 durumlarda b\u00fcy\u00fck bir kolayl\u0131k sa\u011flad\u0131.<\/p>\n<h3>3. Perakende Sekt\u00f6r\u00fcnde M\u00fc\u015fteri Deneyimi \u0130yile\u015ftirme<\/h3>\n<p>B\u00fcy\u00fck bir perakende zinciri, i\u015fitme engelli m\u00fc\u015fterilerine daha iyi hizmet verebilmek amac\u0131yla ma\u011faza i\u00e7i bir mobil uygulama geli\u015ftirdi. Bu uygulama, m\u00fc\u015fterilerin bir \u00fcr\u00fcn\u00fc sorarken veya yard\u0131m isterken yapt\u0131klar\u0131 Libras i\u015faretlerini tan\u0131yarak, ma\u011faza personelinin tabletlerine an\u0131nda metin mesajlar\u0131 g\u00f6nderiyordu. \u00d6rne\u011fin, bir m\u00fc\u015fteri \"Beden\" i\u015faretini yapt\u0131\u011f\u0131nda, uygulama personelin tabletine \"M\u00fc\u015fteri beden hakk\u0131nda bilgi istiyor\" mesaj\u0131n\u0131 iletiyordu.<\/p>\n<p>Bu proje, \u00f6zellikle al\u0131\u015fveri\u015f deneyimini ki\u015fiselle\u015ftirmek ve engelleri azaltmak amac\u0131yla tasarland\u0131. Teachable Machine, \u00fcr\u00fcnlerle ilgili s\u0131k\u00e7a kullan\u0131lan i\u015faretleri ve temel m\u00fc\u015fteri hizmeti ifadelerini tan\u0131mak i\u00e7in e\u011fitildi. Bu sistem, i\u015fitme engelli m\u00fc\u015fterilerin ma\u011fazalarda kendilerini daha rahat ve de\u011ferli hissetmelerini sa\u011flad\u0131, ayn\u0131 zamanda markan\u0131n eri\u015filebilirlik imaj\u0131n\u0131 g\u00fc\u00e7lendirdi.<\/p>\n<p>Bu vaka analizleri, Teachable Machine ve yapay zeka destekli Libras tan\u0131y\u0131c\u0131lar\u0131n sadece teknolojik bir yenilik olmakla kalmay\u0131p, ayn\u0131 zamanda toplumsal fayda sa\u011flayan, ger\u00e7ek d\u00fcnya sorunlar\u0131na pratik \u00e7\u00f6z\u00fcmler sunan g\u00fc\u00e7l\u00fc ara\u00e7lar oldu\u011funu g\u00f6stermektedir. Bu projeler, teknolojinin insan ya\u015fam\u0131n\u0131 nas\u0131l zenginle\u015ftirebilece\u011finin somut \u00f6rnekleridir.<\/p>\n<h2>Sonu\u00e7 ve S\u0131k\u00e7a Sorulan Sorular<\/h2>\n<p>Yapay zeka ve Teachable Machine'in g\u00fcc\u00fcn\u00fc kullanarak bir Libras tan\u0131y\u0131c\u0131 geli\u015ftirmek, ileti\u015fim engellerini a\u015fmak ve daha kapsay\u0131c\u0131 bir toplum yaratmak ad\u0131na at\u0131labilecek de\u011ferli bir ad\u0131md\u0131r. G\u00f6rd\u00fc\u011f\u00fcm\u00fcz gibi, bu s\u00fcre\u00e7 teknik olarak karma\u015f\u0131k gibi g\u00f6r\u00fcnse de, Teachable Machine gibi kullan\u0131c\u0131 dostu ara\u00e7lar sayesinde olduk\u00e7a eri\u015filebilir hale gelmi\u015ftir. Veri toplaman\u0131n \u00f6nemi, model e\u011fitimindeki incelikler ve \u00e7e\u015fitli platformlara entegrasyon yetene\u011fi, bu t\u00fcr projelerin ba\u015far\u0131s\u0131nda kilit rol oynamaktad\u0131r. \u0130ster bir e\u011fitim arac\u0131, ister bir kamu hizmeti kolayla\u015ft\u0131r\u0131c\u0131s\u0131, isterse de perakende sekt\u00f6r\u00fcnde m\u00fc\u015fteri deneyimi iyile\u015ftiricisi olsun, Libras tan\u0131y\u0131c\u0131lar geni\u015f bir uygulama alan\u0131na sahiptir ve gelecekte daha da geli\u015fmeye a\u00e7\u0131kt\u0131r.<\/p>\n<p>Bu teknolojilerin \u00f6n\u00fcndeki en b\u00fcy\u00fck zorluklardan biri, i\u015faret dilinin b\u00f6lgesel farkl\u0131l\u0131klar\u0131 ve her i\u015faretin kendine \u00f6zg\u00fc incelikleri olabilir. Ancak s\u00fcrekli geli\u015fen yapay zeka algoritmalar\u0131 ve artan veri toplama olanaklar\u0131 sayesinde, bu zorluklar\u0131n \u00fcstesinden gelmek m\u00fcmk\u00fcnd\u00fcr. Unutmayal\u0131m ki, her k\u00fc\u00e7\u00fck ad\u0131m, daha eri\u015filebilir ve anlay\u0131\u015fl\u0131 bir d\u00fcnya in\u015fa etme yolunda b\u00fcy\u00fck bir fark yarat\u0131r.<\/p>\n<h3>S\u0131k\u00e7a Sorulan Sorular (SSS)<\/h3>\n<dl>\n<dt>Teachable Machine ile olu\u015fturulan modeller ne kadar do\u011fru \u00e7al\u0131\u015f\u0131r?<\/dt>\n<dd>Modelin do\u011frulu\u011fu, toplad\u0131\u011f\u0131n\u0131z verinin kalitesi ve miktar\u0131yla do\u011frudan orant\u0131l\u0131d\u0131r. Ne kadar \u00e7e\u015fitli, net ve yeterli say\u0131da \u00f6rnekle e\u011fitim yaparsan\u0131z, modeliniz o kadar do\u011fru \u00e7al\u0131\u015f\u0131r. Ortalama olarak, iyi bir veri setiyle %85-95 aras\u0131 do\u011fruluk oranlar\u0131na ula\u015fmak m\u00fcmk\u00fcnd\u00fcr.<\/dd>\n<dt>Libras d\u0131\u015f\u0131ndaki i\u015faret dilleri i\u00e7in de Teachable Machine kullan\u0131labilir mi?<\/dt>\n<dd>Kesinlikle! Teachable Machine, herhangi bir g\u00f6rsel tabanl\u0131 s\u0131n\u0131fland\u0131rma g\u00f6revi i\u00e7in kullan\u0131labilir. Amerikan \u0130\u015faret Dili (ASL), T\u00fcrk \u0130\u015faret Dili (T\u0130D) veya ba\u015fka herhangi bir i\u015faret dili i\u00e7in kendi veri setinizi olu\u015fturup modelinizi e\u011fitebilirsiniz. Yakla\u015f\u0131m ayn\u0131 olacakt\u0131r.<\/dd>\n<dt>Modeli d\u0131\u015fa aktard\u0131ktan sonra internet ba\u011flant\u0131s\u0131 olmadan \u00e7al\u0131\u015f\u0131r m\u0131?<\/dt>\n<dd>Evet, d\u0131\u015fa aktard\u0131\u011f\u0131n\u0131z model (\u00f6zellikle TensorFlow Lite veya TensorFlow.js formatlar\u0131) genellikle bir kez y\u00fcklendikten sonra \u00e7evrimd\u0131\u015f\u0131 \u00e7al\u0131\u015fabilir. Web tabanl\u0131 uygulamalarda, PWA'lar sayesinde model ve di\u011fer kaynaklar \u00f6nbelle\u011fe al\u0131narak \u00e7evrimd\u0131\u015f\u0131 kullan\u0131m sa\u011flanabilir. Mobil uygulamalar i\u00e7in model do\u011frudan cihaza entegre edildi\u011finde zaten \u00e7evrimd\u0131\u015f\u0131 \u00e7al\u0131\u015f\u0131r.<\/dd>\n<dt>Teachable Machine'in alternatifleri var m\u0131?<\/dt>\n<dd>Evet, benzer ama\u00e7lar i\u00e7in Google Cloud AI Platform (AutoML Vision), Microsoft Azure Custom Vision veya Lobe.ai gibi platformlar da mevcuttur. Ancak Teachable Machine, en kolay kullan\u0131ma sahip ve kodlama bilgisi gerektirmeyen ara\u00e7lardan biridir.<\/dd>\n<dt>Bu projeyi daha ileriye ta\u015f\u0131mak i\u00e7in hangi teknolojileri \u00f6\u011frenebilirim?<\/dt>\n<dd>Daha ileri d\u00fczey projeler i\u00e7in Python ile TensorFlow\/Keras, OpenCV (g\u00f6r\u00fcnt\u00fc i\u015fleme), Mediapipe (el\/v\u00fccut poz tespiti) ve mobil uygulama geli\u015ftirme (Android\/iOS) gibi teknolojileri \u00f6\u011frenmek faydal\u0131 olacakt\u0131r. Bu ara\u00e7lar, daha \u00f6zelle\u015ftirilmi\u015f ve karma\u015f\u0131k modeller geli\u015ftirmenize olanak tan\u0131r.<\/dd>\n<\/dl>\n<p><\/body><\/p>\n","protected":false},"excerpt":{"rendered":"\u0130\u015faret diliyle ileti\u015fimi kolayla\u015ft\u0131rmak ve engelleri ortadan kald\u0131rmak m\u0131 istiyorsunuz? Yapay zeka ve Google’\u0131n Teachable Machine arac\u0131yla, Brezilya…","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"csco_page_header_type":"","csco_page_load_nextpost":"","csco_page_subscribe_form":"","csco_page_contact_form":"","footnotes":""},"categories":[1],"tags":[],"class_list":{"0":"post-32060","1":"post","2":"type-post","3":"status-publish","4":"format-standard","6":"category-genel","7":"cs-entry","8":"cs-video-wrap"},"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v20.5 (Yoast SEO v25.3.1) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Yapay Zeka ve Teachable Machine ile Libras Tan\u0131y\u0131c\u0131 Geli\u015ftirme<\/title>\n<meta name=\"description\" content=\"\u0130\u015faret diliyle ileti\u015fimi kolayla\u015ft\u0131rmak ve engelleri ortadan kald\u0131rmak m\u0131 istiyorsunuz? Yapay zeka ve Google'\u0131n Teachable Machine arac\u0131yla, Brezilya \u0130\u015faret Dili (Libras) hareketlerini tan\u0131yabilen bir sistem olu\u015fturmak art\u0131k \u00e7ok daha eri\u015filebilir ve pratik. Bu makale, s\u0131f\u0131rdan ba\u015flayarak kendi Libras tan\u0131y\u0131c\u0131n\u0131z\u0131 nas\u0131l geli\u015ftirece\u011finizi ad\u0131m ad\u0131m a\u00e7\u0131kl\u0131yor.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/\" \/>\n<meta property=\"og:locale\" content=\"tr_TR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Yapay Zeka ve Teachable Machine ile Libras Tan\u0131y\u0131c\u0131 Geli\u015ftirme\" \/>\n<meta property=\"og:description\" content=\"\u0130\u015faret diliyle ileti\u015fimi kolayla\u015ft\u0131rmak ve engelleri ortadan kald\u0131rmak m\u0131 istiyorsunuz? Yapay zeka ve Google'\u0131n Teachable Machine arac\u0131yla, Brezilya \u0130\u015faret Dili (Libras) hareketlerini tan\u0131yabilen bir sistem olu\u015fturmak art\u0131k \u00e7ok daha eri\u015filebilir ve pratik. Bu makale, s\u0131f\u0131rdan ba\u015flayarak kendi Libras tan\u0131y\u0131c\u0131n\u0131z\u0131 nas\u0131l geli\u015ftirece\u011finizi ad\u0131m ad\u0131m a\u00e7\u0131kl\u0131yor.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/\" \/>\n<meta property=\"og:site_name\" content=\"Kodlar\u0131n Gizemli D\u00fcnyas\u0131\" \/>\n<meta property=\"article:published_time\" content=\"2025-10-17T05:01:47+00:00\" \/>\n<meta name=\"author\" content=\"Fatih Soysal\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Yazan:\" \/>\n\t<meta name=\"twitter:data1\" content=\"Fatih Soysal\" \/>\n\t<meta name=\"twitter:label2\" content=\"Tahmini okuma s\u00fcresi\" \/>\n\t<meta name=\"twitter:data2\" content=\"24 dakika\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/\"},\"author\":{\"name\":\"Fatih Soysal\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1\"},\"headline\":\"Yapay Zeka ve Teachable Machine ile Libras Tan\u0131y\u0131c\u0131 Geli\u015ftirme\",\"datePublished\":\"2025-10-17T05:01:47+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/\"},\"wordCount\":4723,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1\"},\"inLanguage\":\"tr\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/#respond\"]}],\"copyrightYear\":\"2025\",\"copyrightHolder\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#organization\"}},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/\",\"url\":\"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/\",\"name\":\"Yapay Zeka ve Teachable Machine ile Libras Tan\u0131y\u0131c\u0131 Geli\u015ftirme\",\"isPartOf\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#website\"},\"datePublished\":\"2025-10-17T05:01:47+00:00\",\"description\":\"\u0130\u015faret diliyle ileti\u015fimi kolayla\u015ft\u0131rmak ve engelleri ortadan kald\u0131rmak m\u0131 istiyorsunuz? Yapay zeka ve Google'\u0131n Teachable Machine arac\u0131yla, Brezilya \u0130\u015faret Dili (Libras) hareketlerini tan\u0131yabilen bir sistem olu\u015fturmak art\u0131k \u00e7ok daha eri\u015filebilir ve pratik. Bu makale, s\u0131f\u0131rdan ba\u015flayarak kendi Libras tan\u0131y\u0131c\u0131n\u0131z\u0131 nas\u0131l geli\u015ftirece\u011finizi ad\u0131m ad\u0131m a\u00e7\u0131kl\u0131yor.\",\"breadcrumb\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/#breadcrumb\"},\"inLanguage\":\"tr\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Anasayfa\",\"item\":\"https:\/\/fatihsoysal.com\/blog\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Yapay Zeka ve Teachable Machine ile Libras Tan\u0131y\u0131c\u0131 Geli\u015ftirme\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#website\",\"url\":\"https:\/\/fatihsoysal.com\/blog\/\",\"name\":\"Fatihsoysal.com\",\"description\":\"Blog - Yaz\u0131l\u0131m D\u00fcnyas\u0131 Tecr\u00fcbelerim\",\"publisher\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/fatihsoysal.com\/blog\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"tr\"},{\"@type\":[\"Person\",\"Organization\"],\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1\",\"name\":\"Fatih Soysal\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"tr\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/fatihsoysal.com\/blog\/wp-content\/uploads\/2024\/04\/cropped-replicate-prediction-3kgg1hgjn5rgp0cf0p5tr0jw7w-1.png\",\"contentUrl\":\"https:\/\/fatihsoysal.com\/blog\/wp-content\/uploads\/2024\/04\/cropped-replicate-prediction-3kgg1hgjn5rgp0cf0p5tr0jw7w-1.png\",\"width\":512,\"height\":512,\"caption\":\"Fatih Soysal\"},\"logo\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/image\/\"},\"description\":\"Kullan\u0131m ve kodlama m\u00fckemmeliyetini odak alan uygulamalar olu\u015fturma deneyimine sahip, profesyonel olarak 15+ y\u0131l \u00fczeri deneyime sahip bir yaz\u0131l\u0131m m\u00fchendisi.\",\"url\":\"https:\/\/fatihsoysal.com\/blog\/author\/fatihsoysal\/\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Yapay Zeka ve Teachable Machine ile Libras Tan\u0131y\u0131c\u0131 Geli\u015ftirme","description":"\u0130\u015faret diliyle ileti\u015fimi kolayla\u015ft\u0131rmak ve engelleri ortadan kald\u0131rmak m\u0131 istiyorsunuz? Yapay zeka ve Google'\u0131n Teachable Machine arac\u0131yla, Brezilya \u0130\u015faret Dili (Libras) hareketlerini tan\u0131yabilen bir sistem olu\u015fturmak art\u0131k \u00e7ok daha eri\u015filebilir ve pratik. Bu makale, s\u0131f\u0131rdan ba\u015flayarak kendi Libras tan\u0131y\u0131c\u0131n\u0131z\u0131 nas\u0131l geli\u015ftirece\u011finizi ad\u0131m ad\u0131m a\u00e7\u0131kl\u0131yor.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/","og_locale":"tr_TR","og_type":"article","og_title":"Yapay Zeka ve Teachable Machine ile Libras Tan\u0131y\u0131c\u0131 Geli\u015ftirme","og_description":"\u0130\u015faret diliyle ileti\u015fimi kolayla\u015ft\u0131rmak ve engelleri ortadan kald\u0131rmak m\u0131 istiyorsunuz? Yapay zeka ve Google'\u0131n Teachable Machine arac\u0131yla, Brezilya \u0130\u015faret Dili (Libras) hareketlerini tan\u0131yabilen bir sistem olu\u015fturmak art\u0131k \u00e7ok daha eri\u015filebilir ve pratik. Bu makale, s\u0131f\u0131rdan ba\u015flayarak kendi Libras tan\u0131y\u0131c\u0131n\u0131z\u0131 nas\u0131l geli\u015ftirece\u011finizi ad\u0131m ad\u0131m a\u00e7\u0131kl\u0131yor.","og_url":"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/","og_site_name":"Kodlar\u0131n Gizemli D\u00fcnyas\u0131","article_published_time":"2025-10-17T05:01:47+00:00","author":"Fatih Soysal","twitter_card":"summary_large_image","twitter_misc":{"Yazan:":"Fatih Soysal","Tahmini okuma s\u00fcresi":"24 dakika"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/#article","isPartOf":{"@id":"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/"},"author":{"name":"Fatih Soysal","@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1"},"headline":"Yapay Zeka ve Teachable Machine ile Libras Tan\u0131y\u0131c\u0131 Geli\u015ftirme","datePublished":"2025-10-17T05:01:47+00:00","mainEntityOfPage":{"@id":"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/"},"wordCount":4723,"commentCount":0,"publisher":{"@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1"},"inLanguage":"tr","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/#respond"]}],"copyrightYear":"2025","copyrightHolder":{"@id":"https:\/\/fatihsoysal.com\/blog\/#organization"}},{"@type":"WebPage","@id":"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/","url":"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/","name":"Yapay Zeka ve Teachable Machine ile Libras Tan\u0131y\u0131c\u0131 Geli\u015ftirme","isPartOf":{"@id":"https:\/\/fatihsoysal.com\/blog\/#website"},"datePublished":"2025-10-17T05:01:47+00:00","description":"\u0130\u015faret diliyle ileti\u015fimi kolayla\u015ft\u0131rmak ve engelleri ortadan kald\u0131rmak m\u0131 istiyorsunuz? Yapay zeka ve Google'\u0131n Teachable Machine arac\u0131yla, Brezilya \u0130\u015faret Dili (Libras) hareketlerini tan\u0131yabilen bir sistem olu\u015fturmak art\u0131k \u00e7ok daha eri\u015filebilir ve pratik. Bu makale, s\u0131f\u0131rdan ba\u015flayarak kendi Libras tan\u0131y\u0131c\u0131n\u0131z\u0131 nas\u0131l geli\u015ftirece\u011finizi ad\u0131m ad\u0131m a\u00e7\u0131kl\u0131yor.","breadcrumb":{"@id":"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/#breadcrumb"},"inLanguage":"tr","potentialAction":[{"@type":"ReadAction","target":["https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/fatihsoysal.com\/blog\/yapay-zeka-ve-teachable-machine-ile-libras-taniyici-gelistirme\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Anasayfa","item":"https:\/\/fatihsoysal.com\/blog\/"},{"@type":"ListItem","position":2,"name":"Yapay Zeka ve Teachable Machine ile Libras Tan\u0131y\u0131c\u0131 Geli\u015ftirme"}]},{"@type":"WebSite","@id":"https:\/\/fatihsoysal.com\/blog\/#website","url":"https:\/\/fatihsoysal.com\/blog\/","name":"Fatihsoysal.com","description":"Blog - Yaz\u0131l\u0131m D\u00fcnyas\u0131 Tecr\u00fcbelerim","publisher":{"@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/fatihsoysal.com\/blog\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"tr"},{"@type":["Person","Organization"],"@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1","name":"Fatih Soysal","image":{"@type":"ImageObject","inLanguage":"tr","@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/image\/","url":"https:\/\/fatihsoysal.com\/blog\/wp-content\/uploads\/2024\/04\/cropped-replicate-prediction-3kgg1hgjn5rgp0cf0p5tr0jw7w-1.png","contentUrl":"https:\/\/fatihsoysal.com\/blog\/wp-content\/uploads\/2024\/04\/cropped-replicate-prediction-3kgg1hgjn5rgp0cf0p5tr0jw7w-1.png","width":512,"height":512,"caption":"Fatih Soysal"},"logo":{"@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/image\/"},"description":"Kullan\u0131m ve kodlama m\u00fckemmeliyetini odak alan uygulamalar olu\u015fturma deneyimine sahip, profesyonel olarak 15+ y\u0131l \u00fczeri deneyime sahip bir yaz\u0131l\u0131m m\u00fchendisi.","url":"https:\/\/fatihsoysal.com\/blog\/author\/fatihsoysal\/"}]}},"yoast_meta":{"yoast_wpseo_title":"","yoast_wpseo_metadesc":"","yoast_wpseo_canonical":""},"amp_enabled":true,"_links":{"self":[{"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/posts\/32060","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/comments?post=32060"}],"version-history":[{"count":0,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/posts\/32060\/revisions"}],"wp:attachment":[{"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/media?parent=32060"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/categories?post=32060"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/tags?post=32060"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}