Image of Penerapan Metode Similarity Search Pada Sistem Music Retrieval Berbasis Konten Audio Musik
Penanda Bagikan

CD Skripsi

Penerapan Metode Similarity Search Pada Sistem Music Retrieval Berbasis Konten Audio Musik

The increasing volume of available digital music data necessitates the development of systems that assist users in discovering songs based on the characteristics of the audio content itself, rather than relying solely on metadata such as titles or artists. This study aims to develop a Content-Based Music Retrieval (CBMR) system that enables music search through the audio features of songs, allowing the construction of unique representations for each track. Segments of audio are converted into image form containing feature scores. An image-based similarity matching algorithm is employed to compare the user's input audio with a music database, enabling the system to identify songs with similar melodic characteristics. In this research, the development process of the CBMR system involves music data collection, feature extraction, feature database construction, and similarity-based image retrieval using the VGG-19 model. A total of 60 music samples were used, from which frequency and amplitude features were extracted. The VGG-19 architecture processes and classifies features such that the audio segments—converted into images—pass through three fully connected (FC) layers, with FC1 storing complex features. The results show that the top five retrievals achieved a highest similarity score of 41.93% for frequency-based similarity and 30.62% for amplitude-based similarity. It is important to note that the system's performance may be influenced by the quality of the input audio image. With further development, this system has the potential to make a significant contribution to the field of digital music retrieval and can be implemented in music streaming services or other music platforms to enhance the user experience.

Key Words: Content-Based Music Retrieval, Deep Image Search, music information system, music similarities, VGG-19.

Ketersediaan
#
Perpustakaan Universitas Riau 1903124580
1903124580
Tersedia
Informasi Detail
Judul Seri
-
No. Panggil
1903124580
Penerbit
Pekanbaru : Universitas Riau FMIPA Sistem Informasi.,
Deskripsi Fisik
-
Bahasa
Indonesia
ISBN/ISSN
-
Klasifikasi
1903124580
Tipe Isi
-
Tipe Media
-
Tipe Pembawa
-
Edisi
-
Subjek
SISTEM INFORMASI
Info Detail Spesifik
-
Pernyataan Tanggungjawab
Versi lain/terkait

Tidak tersedia versi lain

Lampiran Berkas
Komentar
Anda harus masuk sebelum memberikan komentar