Album Cover
Album cover design, crucial for attracting listeners and reflecting musical identity, is increasingly being automated through AI. Current research focuses on developing efficient and accessible tools using various deep learning architectures, including GANs, diffusion models, and genetic algorithms, to generate visually appealing covers based on audio features, lyrics, or genre tags. These advancements aim to empower musicians with limited resources and streamline the design process, improving the overall discoverability and presentation of music. The resulting algorithms are evaluated based on human perception of the generated art's coherence with the music and its aesthetic appeal.
Papers
August 3, 2024
January 13, 2023
July 14, 2022