Sliding Alignment Windows for Real-Time Crowd Captioning
Mohammad Kazemi, Rahman Lavaee, Iftekhar Naim, Daniel Gildea · 2014
The primary way of providing real-time speech to text captioning for hard of hear-ing people is to employ expensive profes-sional stenographers who can type as fast as natural speaking rates. Recent work has shown that a feasible alternative is to com-bine the partial captions of ordinary typ-ists, each of whom is able to type only part of what they hear. In this paper, we extend the state of the art fixed-window alignment algorithm (Naim et al., 2013) for combining the individual captions into a final output sequence. Our method per-forms alignment on a sliding window of the input sequences, drastically reducing both the number of errors and the latency of the system to the end user over the pre-viously published approaches. 1