Many real world problems quirea degree of flexibility that is to achieve using hand algorithms. One such domain is vision-based autonomous driving. In this task, the dual challenges of a constantly changing environment coupled with a real processing constrain the flexibility and of a machine learning system essential. This describes just such a learning system, called (Autonomous Land Vehicle In a Neural Network). It presents the neural network architecture and training techniques that allow to drive in a variety of including singlelane paved and unpaved roads, multilane lined and unlined roads, and obstacle-ridden on- and road environments, at speeds of up to 55 miles hour.

Many real world problems requireadegree of #exibility that is di#- cult to achieve using hand programmed algorithms. One such domain is vision-based autonomous driving. In this task, the dual challenges of a constantly changing environment coupled with a real time processing constrain make the #exibility and e#ciency of a machine learning system essential.

this paper, semi-continuous HMMs (SCHMMs) (Bellagarda & Nahamoo 89; Huang & Jack 89) and continuous densities HMMs (CDHMMs) will be considered in conjunction with networks trained with the generalized delta rule (Rumelhart et al 86). It will be shown how to perform a joint global optimi ation of both the ANN and the HMM parameter estimation. In the proposed algorithm, the gradient of the optimization criterion with respect to the transformed observations is computed for the HMM system. The HMM can be trained with traditional methods (Rabiner 89) with which the gradient of an optimization criterion is computed. This gradient is sent to the ANN for the estimation of the weight associated to each connection of the network. No assumption need to be made or constraints imposed on the network outputs, except that the network output distribution should be modeled by a mixture of multivariate gaussians. Since training of HMMs is usually much faster than ANN training, we consider how to initialize the ANN in order to start from parameter values that are not too far from those obtained after training. Multiple ANNs are combined and an incremental design method is described in which specialized networks are integrated to the recognition system in order to improve its performance. Relate or Interesting papers have been published recently, describing attempts at com-

We study different approaches for text-independent talker identification. We first present and compare three different systems which are based respectively on Connectionist Models, Hidden Markov Models and Multivariate Auto-regressive Models. These three techniques have very different characteristics and capabilities, we compare their performances on the TIMIT database and discuss their respective merits. The three models having been used only very recently for talker recognition, this study is thus exploratory. It is the first step for building more complex systems where the different techniques cooperate together in order to take advantage of their respective merits. We discuss different possibilities for this cooperation . Keywords: Talker identification, Shift Time Delay Neural Networks, Hidden Markov Models, Multivariate AutoRegressive Models, Vectorial Log Likelihood Ratio. 1. Introduction Talker recognition usually refers to three related problems which are the verification o...

Contents 1 Introduction 1 2 The Human Speech 3 2.1 Phonemes : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 3 2.1.1 Other Speech Units : : : : : : : : : : : : : : : : : : : : : 4 2.2 Kinds of Phonemes : : : : : : : : : : : : : : : : : : : : : : : : : : 5 2.2.1 Consonants : : : : : : : : : : : : : : : : : : : : : : : : : : 5 2.2.1.1 Voicing : : : : : : : : : : : : : : : : : : : : : : : 6 2.2.1.2 Place of Articulation : : : : : : : : : : : : : : : 6 2.2.1.3 Manner of Articulation : : : : : : : : : : : : : : 7 2.2.2 Vowels : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 8 2.2.3 Diphthongs : : : : : : : : : : : : : : : : : : : : : : : : : : 8 2.3 Formants : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 9 3 A Signal Processing View of the Human Speech 10 3.1 Def

In this project the use of artificial neural networks for autonomous driving tasks is investigated, especially for obstacle avoidance and road following. We have analysed neural networks with input from various sensor types like a single camera, stereo vision, depth information and linear cameras. During the investigation the resulting driving behaviour of the autonomous mobile agent is tested in a virtual environment. We found that artificial neural networks with input from single sensors result in the best driving behaviour when using depth information. We discovered that combining different sensor inputs with an artificial neural network can generate a better fitting steering output for autonomous mobile agent than with the information of only one sensor. Furthermore, we present interesting