Automatic Chat Transcription on a Firefighter TETRA Broadcast Channel

For a reliable keyword extraction on firefighter radio communication, a strong automatic speech recognition system is needed. However, real-life data poses several challenges like a distorted voice signal, background noise and several different speakers. In this paper, we review our experiences with the PRONTO corpus, which has been recorded during a firefighting exercise. Then, we proceed to present the benchmarks of our chat transcription system in terms of word error rate. Since a large amount of sentences in public safety communication share similar patterns, we also analyse the impact of the sentence complexity on the system performance, and further investigate how much training utterances are needed for a reliable speaker identification in this setting.