An approach to optimize inference of the DIART speaker diarization pipeline
This work addresses low-latency requirements for real-time transcription in speaker diarization, but it is incremental as it applies standard optimization techniques to an existing pipeline.
The paper tackled optimizing inference latency in the DIART online speaker diarization pipeline by applying methods like knowledge distillation, pruning, quantization, and layer fusion to its embedding model, finding that quantization and layer fusion reduced latency without accuracy loss, while knowledge distillation improved latency but hurt accuracy, and pruning had no effect.
Speaker diarization answers the question "who spoke when" for an audio file. In some diarization scenarios, low latency is required for transcription. Speaker diarization with low latency is referred to as online speaker diarization. The DIART pipeline is an online speaker diarization system. It consists of a segmentation and an embedding model. The embedding model has the largest share of the overall latency. The aim of this paper is to optimize the inference latency of the DIART pipeline. Different inference optimization methods such as knowledge distilation, pruning, quantization and layer fusion are applied to the embedding model of the pipeline. It turns out that knowledge distillation optimizes the latency, but has a negative effect on the accuracy. Quantization and layer fusion also have a positive influence on the latency without worsening the accuracy. Pruning, on the other hand, does not improve latency.