“I believe that training methods for recursive neural networks will be some of the most interesting future research in the field. We've seen some pretty revolutionary techniques come out for feedforward models in the last 30 years, but with recurrent/recursive networks it's often difficult to see what's going on inside the network, let alone properly train…”
Speech Recognition with Deep Recurrent Neural Networks
Recurrent neural networks (RNNs) are a powerful model for sequential data. End-to-end training methods such as Connectionist Temporal Classification make it possible to train RNNs for sequence labelling problems where the input-output alignment is unknown. The combination of these methods with the Long Short-term Memory RNN architecture has proved particularly fruitful, delivering state-of-the-art results in cursive handwriting recognition. However RNN performance in speech recognition has so far been disappointing, with better results returned by deep feedforward networks. This paper investigates \emph{deep recurrent neural networks}, which combine the multiple levels of representation that have proved so effective in deep networks with the flexible use of long range context that empowers RNNs. When trained end-to-end with suitable regularisation, we find that deep Long Short-term Memory RNNs achieve a test set error of 17.7% on the TIMIT phoneme recognition benchmark, which to our knowledge is the best recorded score.
候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。
HN 书评
推荐在前,同类按时间 · 每篇最多保留 20 条,统计用全部提及“Yup. Copying my post from reddit: Seems like the title should really be criticisms of STT groups in the industry, work from academic groups is ignored. The original article was kind of silly anyways, it was all engineering work, very little research. Engineering work is important, but you're not going to get an image net moment in ASR by twiddling with …”