Research on the Collaborative Mechanism of Music Melody Generation and Manual Revision Based on AI Large Models

Abstract

With the deep penetration of artificial intelligence technology in the field of cultural creation, music melody generation is gradually
transitioning from being human-dominated to a human-computer collaborative paradigm. AI large models, leveraging their ability to learn
from massive amounts of data, can quickly generate a wide variety of melody materials, but they often fall into the problem of homogenization due to a lack of emotional warmth and artistic originality; traditional human creation can accurately grasp artistic expression but is limited by efficiency bottlenecks and inspiration constraints. The divide between the two makes it difficult for music creation to simultaneously
achieve efficiency, diversity, and artistic depth. Against this background, exploring a reasonable collaborative path between AI generation and
human revision becomes key to breaking through creative bottlenecks. This paper focuses on the collaborative mechanism of AI large modelbased music melody generation and human revision and conducts a systematic study on this aspect.