Several methods of combination of Multilayer Perceptrons (MLPs) for handwritten character recognition are presented and discussed. Recognition tests have shown that cooperation of neural networks using different featu...
详细信息
Several methods of combination of Multilayer Perceptrons (MLPs) for handwritten character recognition are presented and discussed. Recognition tests have shown that cooperation of neural networks using different features vectors can reduce significantly the overall misclassification error rate. The final recognition system consists of a cascade association of small MLPs, which allows minimization of the overall recognition time while retaining a high recognition rate. This system appears to be 50% faster than the best of the individual MLPs, while offering a recognition rate of 99.8% on unconstrained digits extracted from the NIST 3 database.
In this paper, we present the first step of a project that is able to perform both speech and singing synthesis controlled in real-time. Our aim is to develop a flexible application allowing performers to produce comp...
详细信息
In this paper, we present the first step of a project that is able to perform both speech and singing synthesis controlled in real-time. Our aim is to develop a flexible application allowing performers to produce complex and versatile singing - as well as speech - articulations. Thus, we have adapted an existing speech synthesizer, the MBROLA software, to real-time singing constraints. The work presented in this paper concerns the integration of the MBROLA speech synthesizer into the Max/MSP real-time environment through the development of an external object. We present real-time control functions of the object and discuss its limitations. We further discuss about perspectives in the development of MBROLA-based computer music applications.
Dealing with noise deteriorating the speech is still a major problem for automatic speech recognition. An interesting approach to tackle this problem consists of using multi-task learning. In this case, an efficient a...
详细信息
暂无评论