Reasoning LLM Improves Speaker Recognition in Long-form TV Dramas

Long-form TV dramas present a formidable challenge for comprehensive video understanding, where deciphering complex storyline often relies on $\mathbf{\text{speaker}}\mathbf{\text{recognition}}$, the task of accurately attributing each spoken utterance to its respective character. In this paper, we advance this field through two primary contributions. (1) We introduce $\mathbf{\text{DramaSR-532K}}$, a large-scale benchmark comprising 532K annotated dialogue lines across more than 900 unique characters, necessitating the integration of auditory, linguistic, and visual cues for speaker recognition. (2) We propose $\mathbf{\text{DramaSR-LRM}}$, a robust approach built upon a large reasoning model (LRM). DramaSR-LRM is designed to autonomously aggregate contextual evidence via multimodal tool-use, synthesizing diverse inputs to achieve high-fidelity attribution. Experimental results demonstrate that DramaSR-LRM significantly outperforms existing baselines, particularly on short utterances where acoustic biometrics are inherently unreliable. $\mathit{\text{All}}\mathit{\text{the}}\mathit{\text{data}}\mathit{\text{and}}\mathit{\text{code}}\mathit{\text{will}}\mathit{\text{be}}\mathit{\text{made}}\mathit{\text{publicly}}\mathit{\text{available}}\mathit{\text{at}}\mathit{\text{the}}\mathit{\text{project}}\mathit{\text{page:}}\mathit{\text{https://www.github.com/198808xc/DramaSR-LRM.}}$