Abstract

We introduce a direct model for speech recognition that assumes an unstructured, i.e., flat text output. The flat model allows us to model arbitrary attributes and dependencies of the output. This is different from the HMMs typically used for speech recognition. This conventional modeling approach is based on sequential data and makes rigid assumptions on the dependencies. HMMs have proven to be convenient and appropriate for large vocabulary continuous speech recognition. Our task under consideration, however, is the Windows Live Search for Mobile (WLS4M) task [1]. This is a cellphone application that allows users to interact with web-based information portals. In particular, the set of valid outputs can be considered discrete and finite (although probably large, i.e., unseen events are an issue). Hence, a flat direct model lends itself to this task, making the adding of different knowledge sources and dependencies straightforward and cheap. Using e.g. HMM posterior, m-gram, and spotter features, significant improvements over the conventional HMM system were observed.