Wenyuan Li Guang Li 0008 Keisuke Maeda Takahiro Ogawa 0001 Miki Haseyama Decoupled Audio-Visual Dataset Distillation. 2025 November abs/2511.17890 CoRR https://doi.org/10.48550/arXiv.2511.17890 db/journals/corr/corr2511.html#abs-2511-17890 streams/journals/corr