Triangulated Transcript Pipeline for Multi-Source Incident Data

Create a pipeline that fuses department captions, ASR hypotheses from each source, and available court/911 transcripts to produce consensus transcripts with provenance and confidence scores. This pipeline should surface overlaps and contradictions across sources to improve transcript accuracy and reliability.

References:

1. Constructing Datasets From Public Police Body Camera Footage. Jamie Rosas-Smith, Martijn Bartelds, Ruizhe Huang, Leibny Paola García-Perera, Karen Livescu, Dan Jurafsky, Anjalie Field (2025). IEEE International Conference on Acoustics, Speech, and Signal Processing.
2. Body camera footage as data: Using natural language processing to monitor policing at scale & in depth. Nicholas P. Camp, Rob Voigt (2024). Behavioral Science & Policy.
3. Developing Speech Processing Pipelines for Police Accountability. Anjalie Field, Prateek Verma, Nay San, J. Eberhardt, Dan Jurafsky (2023). Interspeech.