Artificial Creatures Lab

 Projects

Computer Engineering Department - Sharif University of Technology  

 Home | Projects | Members | Publications

 Zamin Artificial Life Model    Biped Robot

 In this project, we are creating an artificial world, suitable for training artificial agents and perform cognitive studies.

 

 This project deals with developing a useful robust controller for a biped robot to make it walk in an unstructured environment.

     
 Artificial Emotions    Speech Recognition

 This project is about to make artificial emotions and emotional control. Due to the tight cooperation between this project and project Zamin, refer to Zamin for more details.

 

 Current speech recognition systems usually fail to work in real environments. We believe that a fuzzy method must be able to solve the problem....

     
 Lamp Post Climbing Robot   Exoskeleton

 A project to design and build a robot able to climb lamp posts and replace the lamps, automatically

   A project to implement and control an exoskeleton.
     
 An Underwater Robot   Fuzzy Population Based Optimization

 The goal of the project is to design and implement an stability controller for a remotely operated underwater vehicle.

   The aim of this project is to design  new population-based approach for optimization problems where the search space is traversed via fuzzy rule-based approach.. here crossover and mutation operators are left out!
     
     
 Fuzzy Speech Recognition   Learning in Fuzzy Finite State Automata

 This project is about isolated speech  recognition. It is supposed to use different methods of learning to adjust the parameters of a special fuzzy model in order to recognize speech words.

 

 In this project we want to find a mapping between the concept of state in FSM and granule and using this mapping for learning new experiences. And refining the FSM or adding a new state to it.

     
Fuzzy Pattern Recognition    ACMoD, ACL Modular Robot

 This project is about using a new fuzzy pattern recognition approach to recognize online handwrittens.

 

 A new modular robot designed, which can move individually. These modules can attach together to form more complex structures for different terrains.

     
 Fuzzy Control of Quadrotor    Clustering, Clustering & Classification Dataset  

 In this project fuzzy control algorithms have been employed to control the attitude of a Quadrotor.

 

 In this project a new Clustering Algorithm based on the Active Learning Method, is developed.

     

 

 

   
Former Projects:    
 Combinatorial Optimization    Roboshark, A Billiard Machine

 In this project, Active Learning Method and Reinforcement Learning are used to guide local search algorithms to efficiently solve hard Combinatorial Optimization Problems.

 

 The goal of the project is to design and implement a gantry robot (a robot which its arm's position is determined by moving on two axis, X and Y) with ability to play snooker with an acceptable skill.