RituScape: Automatic Season-Based Virtual Environment Generation from Indian Classical Ragas
摘要
Indian classical ragas are intrinsically linked to particular emotions, times of day, and seasons, creating a rich multisensory experience. We present RituScape, a novel system that automatically adapts virtual environments according to each raga’s seasonal context. Our approach utilizes a curated dataset of ragas spanning six seasonal classes (Ritus), alongside a CNN-based spectrogram analysis that achieved an 88.4% classification accuracy in mapping ragas to their associated seasons. In a mixed-methods user study, RituScape received a mean overall satisfaction score of 4.2 out of 5, indicating heightened immersion and a stronger emotional connection when experiencing ragas in these seasonally adapted environments. This demonstrates the system’s potential for preserving cultural heritage and expanding interactive media applications, particularly in music-based therapeutic and educational contexts. Demo and additional materials are available on GitHub. ( https://github.com/Anonymous1-star-jpg/Anonymous2.git ).