@misc{indiciae47e7bdf09e36, title = {From Who Said What to Who They Are: Modular Training-free Identity-Aware LLM Refinement of Speaker Diarization}, author = {Yu-Wen Chen and William Ho and Maxim Topaz and Julia Hirschberg and Zoran Kostic}, year = {2026}, url = {https://arxiv.org/abs/2509.15082}, note = {Source identifier: 2509.15082} }