My use case is to repeat a sentence after making a mistake while recording, the first sentence always gets edited out and replaced with the mistake-free second take. Being able to detect the same sentence back-to-back would save a lot of time in editing.