Saman Harati Zadeh

Sharif University of Technology, Teheran, Tehrān, Iran

Are you Saman Harati Zadeh?

Claim your profile

Publications (12)3.67 Total impact

  • Source
    Saman Harati Zadeh · Saeed Bagheri Shouraki · Ramin Halavati
    [Show abstract] [Hide abstract]
    ABSTRACT: There are several approaches to emotions in AI, most of which are inspired by human emotional states and their arousal mechanisms. These approaches usually use high-level models of human emotions that are too complex to be directly applicable in simple artificial systems. It seems that a new approach to emotions, based on their functional role in information processing in mind, can help us to construct models of emotions that are both valid and simple. In this paper, we will try to present a model of emotions based on their role in controlling the attention. We will evaluate the performance of the model and show how it can be affected by some structural and environmental factors.
    Preview · Conference Paper · Jun 2008
  • Source
    Saman Harati Zadeh · Saeed Bagheri Shouraki · Ramin Halavati
    [Show abstract] [Hide abstract]
    ABSTRACT: To produce emotional artificial systems in AI domain, usually a subset of human emotional states are imported to the target domain and the major differences between natural and artificial domains are often ignored. In this paper we will discuss about why such an approach is not useful for all possible applications of emotions and we will show how it is necessary and possible to produce artificial emotion systems based on the target systems goals, abilities and needs.
    Preview · Conference Paper · Jan 2008
  • Ramin Halavati · Saeed Bagheri Shouraki · Saman Harati Zadeh
    [Show abstract] [Hide abstract]
    ABSTRACT: Recognition of human speech has long been a hot topic among artificial intelligence and signal processing researches. Most of current policies for this subject are based on extraction of precise features of voice signal and trying to make most out of them by heavy computations. But this focus on signal details has resulted in too much sensitivity to noise and as a result, the necessity of complex noise detection and removal algorithms, which composes a trade-off between fast or noise robust recognition.This paper presents a novel approach to speech recognition using fuzzy modeling and decision making that ignores noise instead of its detection and removal. To do so, the speech spectrogram is converted into a fuzzy linguistic description and this description is used instead of precise acoustic features. During the training period, a genetic algorithm finds appropriate definitions for phonemes, and when these definitions are ready, a simple novel operator consisting of low cost functions such as Max, Min, and Average makes the recognition. The approach is tested on a standard speech database and is compared with Hidden Markov model recognition system with MFCC features as a widely used speech recognition approach.
    No preview · Article · Jun 2007 · Applied Soft Computing
  • Ramin Halavati · Saeed Bagheri Shouraki · Saman Harati Zadeh

    No preview · Article · Jan 2007
  • Saman Harati Zadeh · Saeed Bagheri Shouraki · Ramin Halavati
    [Show abstract] [Hide abstract]
    ABSTRACT: Emotions are the subject of study in many research areas, including psychology, physiology and arti ficial intelligence. However, there is not yet a general computational definition for emotions and many of the related works in artificial intelligence are based on uncertain assumptions about the origin, fea tures, and applications of emotions. Therefore, we direct our studies toward achieving a better understanding of emotions and generating a computational model that could be a base for further research in the field. As a first step, we present the results of our study into the possible role of emotions in men tal resource management. To do so, we will introduce a model of emotions based on managing mental resources. We have applied it to a general model of mind and the whole model has been implemented in an artificial life simulation environment called Zamin to evaluate its ability to produce emotional behavior and to make better decisions. The results, as discussed in this article, show that a resource management mechanism could be the source of emotion generation in the mind.
    No preview · Article · Dec 2006 · Adaptive Behavior
  • Source
    R. Halavati · S.B. Shouraki · S.H. Zadeh · P. Ziaie · C. Lucas
    [Show abstract] [Hide abstract]
    ABSTRACT: Zamin artificial life model is designed to be a general purpose environment for researches on evolution of learning methods, living strategies and complex behaviors and is used in several studies thus far. As a main target for Zamin's design has been its expandability and ease of problem definition, a new agent based structure for this artificial world is introduced in this paper, which is believed to be much easier to use and extend. In this new model, all control and world running processes are done by agents. Therefore, any change in world processes does not require recoding the main engine and can be done just by altering the behavior of one or some agents. To have an easier interface for the design of new organisms, all creatures' communications with the world level is done through a common message map, thus, a designer just needs to code the required parts and append them to the main system to process the necessary messages. And at last, extending the model can be done with much less effort, as it can be done easily by creating new agents that handle the new tasks. This model is implemented and some coding differences with previous model are presented.
    Preview · Conference Paper · Jan 2005
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Packing problems arise in a wide variety of applica-tion areas. The basic problem is that of determining an efficient ar-rangement of different objects in a region without any overlap and with minimal wasted gap between shapes. This paper presents a novel population based approach for optimizing arrangement of ir-regular shapes. In this approach, each shape is coded as an agent and the agents' reproductions and grouping policies results in arrange-ments of the objects in positions with least wasted area between them. The approach is implemented in an application for cutting sheets and test results on several problems from literature are pre-sented.
    Preview · Article · Jan 2005
  • R. Halavati · S.H. Zadeh · S.B. Shouraki
    [Show abstract] [Hide abstract]
    ABSTRACT: Zamin, which is a high level artificial life environment have been successfully used as a test bed for a number of cognitive and AI studies. Here we have tried to test the evolution of a pleasure computing mechanism in Zamin's artificial creatures and have extended their mental capabilities to cover uncertainty in action selection mechanism. The results show some improvements in both genetic evolution process and learning capabilities. More specifically, we have evolved an internal pleasure system in Zamin creatures for the first time, quite unsupervised. In addition creatures could learn much more efficient behavioral patterns than what they could before.
    No preview · Article · Jan 2004
  • Source
    SAMAN HARATI ZADEH · ABOLFAZL KEIGHOBADI LAMJIRI · CARO LUCAS
    [Show abstract] [Hide abstract]
    ABSTRACT: Zamin, is a general purpose artificial environment suitable for cognitive studies. This environment, which presents high-level agents in a rather complex environment, has been successfully used as a test bed for some AI researches, like cooperative learning. Although it still has some deficiencies, it has many useful features and in addition, it can be enhanced with some proposed features to cover more capabilities. We have prepare a survey on some recent researches in the area of Artificial Life to show the advantages of this environment in comparison with nowadays special purpose as well as general purpose simulated world test-beds. In addition we have tried to highlight its disadvantages that may be fixed by researchers that are interested in this subject. Source code and documentation of both versions of Zamin (with normal as well as fuzzy learners) are publicly available on the web.
    Preview · Article · Jan 2004
  • S.H. Zadeh · R. Halavati · S.B. Shouraki
    [Show abstract] [Hide abstract]
    ABSTRACT: Zamin, is an Artificial Life ecosystem, suitable for cognitive studies such as evaluation of different thinking and acting methods. Up to now, simple models for decision making, generalization and abstraction, environment perception and pleasure computation is developed and tested in Zamin. In this paper, we have promoted the two layered decision-making system of Zamin creatures, the Aryos, to a three layered emotional decision making mechanism, compared the learning and coping with predators capabilities of the new creatures with the old ones and shown that they can defeat the previous, emotion-less creatures.
    No preview · Conference Paper · Jan 2004
  • Source
    SAMAN HARATI ZADEH · RAMIN HALAVATI · SAEED BAGHERI SHOURAKI
    [Show abstract] [Hide abstract]
    ABSTRACT: Zamin that has been introduced as a high level Artificial Life environment has a variety of good features that make it a general purpose ALife test bed in which different aspects of real life may be modeled and studied. In this paper we will first describe the general structures of Zamin and living mechanisms of its creatures and then we will discuss about some of our successful studies in Zamin test bed to show the ways in which this simulated life environment can be used for cognitive researches.
    Preview · Article ·
  • Source
    Ramin Halavati · Saeed Bagheri Shouraki · Saman Harati Zadeh · Caro Lucas
    [Show abstract] [Hide abstract]
    ABSTRACT: Artificial Life can be used as an agent training approach in large state spaces. This paper presents an artificial life method to increase the training speed of some speech recognizer agents which where previously trained by genetic algorithms. Using this approach, vertical training (genetic mutations and selection) is combined with horizontal training (individual learning through reinforcement learning) and results in a much faster evolution than simple genetic algorithm. The approach is tested and a comparison with GA cases on a standard speech data base is presented.
    Preview · Article ·