Hasil Pencarian  ::  Simpan CSV :: Kembali

"Proses desain dari sebuah website dimulai dengan membuat wireframe untuk masing-masing halaman web dengan menggunakan gambar tangan (hand-drawn wireframe) di kertas atau menggunakan peralatan desain grafis yang dikhususkan untuk membuat wireframe. Wireframe kemudian akan di konversikan kedalam bentuk HTML struktural ataupun kedalam bentuk markup code lain secara manual oleh software engineer. Proses ini biasanya terjadi berulang kali hingga beberapa tahap yang cukup lama sampai akhirnya code tersebut berhasil dibuat. Oleh karena itu dibutuhkan sistem yang dapat melakukan proses konversi hand-drawn web wireframe kedalam bentuk HTML dasar secara otomatis dengan menerapkan pendekatan deep learning. Sistem ini dibangun dengan menggunakan metode image captioning yang didalamnya terdiri dari CNN encoder, LSTM decoder, dan compiler. Sistem ini di uji dengan beberapa hand-drawn web wireframe dengan kostumisasi bentuk, yang diambil secara langsung, untuk melihat tingkat akurasi deteksi sketsa dan tingkat akurasi konversi berdasarkan masing-masing layout dan element HTML. Nilai akurasi deteksi dan akurasi konversi ini di ukur menggunakan performance metric dan BLEU score. Dari proses pengujian didapatkan nilai akurasi sistem sebesar 83% dan BLEU score sebesar 0.93.

---

The design process of a website begins with the creation of a hand-drawn wireframe for each web page or using a wirefme specified graphic design software. Afterwards, the created wireframe will be converted into a HTML structure or into other markup codes manually by a software engineer. This process typically occurs repeatedly with steps that take up a considerable amount of time until finally the code is successfully created. Therefore, a system with deep learning as an approach that could perform a conversion from a hand-drawn web wireframe into basic HTML structure automatically is needed. This system was built with image captioning method which consists of a CNN encoder, a LSTM decoder, and a compiler. This system was tested through several hand-drawn web wireframes with shape costumisation which were taken directly to see the accuracy of sketch detection and the accuracy of the convertion based on each layout and HTML element. The detection accuracy score and the convertion accuracy score was measured with a performance metric and BLEU score. From the testing process, the system's accuracy of 83% was obtained with a BLEU score of 0.93."

"This handbook is a comprehensive and up-to-date resource covering the booming field of Audiovisual Translation (AVT) and Media Accessibility (MA). Bringing together an international team of renowned scholars in the field of translation studies, the handbook surveys the state of the discipline, consolidates existing knowledge, explores avenues for future research and development, and also examines methodological and ethical concerns. This handbook will be a valuable resource for advanced undergraduate and postgraduate students, early-stage researchers but also experienced scholars working in translation studies, communication studies, media studies, linguistics, cultural studies and foreign language education."