TY  - RPRT
TI  - SPHERE: Automatic Music Upmixing via Audio Language Model Post-Training with Spatial Heuristic Rewards
AU  - Zixun Guo
AU  - Calvin Murdock
AU  - Sanjeel Parekh
AU  - W Owen Brimijoin
AU  - Simon Dixon
AU  - Joshua Reiss
AU  - Ishwarya Ananthabhotla
PY  - 2026
UR  - https://arxiv.org/abs/2608.30559
ID  - 2608.30559
ER  -