Sintesa Ekspresi Emosi Robot Humanoid Multimodal

Hermawanto, Fajar (2015) Sintesa Ekspresi Emosi Robot Humanoid Multimodal. Masters thesis, Institut Teknologi Sepuluh Nopember.

[img]
Preview
Text
2212206014-Master Thesis.pdf - Published Version

Download (3MB) | Preview

Abstract

Manusia berinteraksi melalui tiga saluran utama yaitu visual ( penglihatan ) haptic ( peraba ), audio ( suara ). Dalam berinteraksi manusia mengungkapkan emosi dalam bentuk ekspresi. Robot humanoid diciptakan agar berprilaku seperti manusia. Interaksi antara robot dengan manusia masih terdapat keterbatasan , yaitu robot belum mampu mengenali ekspresi emosi manusia. Robot juga belum mampu mengekspresikan emosi yang telah dikenali sehingga dapat diterima oleh manusia. Dengan adanya keterbatasan tersebut maka perlu diteliti agar robot mampu mengenali ekspresi emosi dan mengekspresikan emosi yang telah dikenali tersebut. Pengenalan ekspresi emosi manusia salah satunya melalui klasifikasi ekspresi emosi wajah. Pada saat ini klasifikasi ekspresi emosi wajah dapat dilakukan oleh komputer, salah satunya menggunakan program emotionviewer yang terdapat pada Intel RealSense SDK. Program emotion viewer dapat mengklasifikasi ekspresi emosi wajah senang, sedih, marah, takut, terkejut, dan jijik. Program tersebut dapat digunakan pada robot sebagai pengindraan ekspresi emosi wajah pengguna. Setelah ekspresi emosi dikenali robot mengekspresikan emosi sehingga dapat diterima oleh pengguna. sintesa ekspresi robot dapat melalui visual ( gerakan, warna ), audio ( nada ). Pengujian sintesa ekspresi emosi robot menggunakan kuisioner yang dilakukan secara online. Berdasarkan data hasil kuisioner penelitian menunjukkan hasil yang baik dengan tingkat akurasi sebesar 80%. Robot mampu berprilaku seperti manusia yaitu mampu mengenali ekspresi emosi wajah pengguna dan mensintesakan ekspresi emosi yang telah dikenali. ======================================================================================================== Humans interact through three main channels, namely visual (sight) haptic (touch), audio (sound). In humans interact expressing emotion in the form of expression. Humanoid robot created in order to behave like a human. The interaction between robots with humans there are limitations, the robot is not able to recognize the expression of human emotions. The robot also has not been able to express the emotions that have been identified that can be accepted by humans. With these limitations, the need to be investigated so that the robot is able to recognize the expression of emotions and express the emotions that have been identified. Introduction to the expression of human emotion one through classification facial emotional expressions. At this time the classification of facial emotional expressions can be done by computer, one of them using the programs contained in the Intel emotionviewer RealSense SDK. Emotion viewer program can classify facial emotional expressions happy, sad, angry, scared, shocked, and disgusted. The program can be used on the robot as an expression of emotion sensing the user's face. After the robot recognized emotional expressions that express emotions can be accepted by the user. synthesis robot can through visual expression (movement, color), audio (tone). Testing synthesis robot emotion expression using questionnaires conducted online. Based on data from a questionnaire study showed good results with an accuracy rate of 80%. The robot is able to behave like humans are able to recognize facial expressions of emotion users and synthesizing emotional expression has been identified

Item Type: Thesis (Masters)
Additional Information: RTE 629.892 Her s
Uncontrolled Keywords: Robot humanoid, ekspresi, emosi, wajah, emotionviewer, Intel RealSense SDK, sintesa, gerakan, warna, nada
Subjects: T Technology > TK Electrical engineering. Electronics Nuclear engineering
Divisions: Faculty of Industrial Technology > Electrical Engineering > (S2) Master Theses
Depositing User: Mr. Tondo Indra Nyata
Date Deposited: 26 Mar 2018 08:40
Last Modified: 24 Aug 2018 03:09
URI: http://repository.its.ac.id/id/eprint/51639

Actions (login required)

View Item View Item