Machine Learning and Robotics Department, the School of Electrical and Computer Engineering, University of Tehran
Projects’ Supervisor: Prof. Hadi Moradi
Curator: Martub Projects
Recently, the human brain has been the subject of much scholarly research. Everyone wants to know how the brain works and how we can build a machine with similar abilities. Meantime, researchers are seeking for the reasons of cognitive disorders of the brain, trying to find solutions to cure/rehabilitate those with the problem. Thus, there has been considerable development in the field of cognitive sciences. The Faculty of Electrical and Computer Engineering of the University of Tehran, particularly Robotics and Machine Learning Department, has made great effort in cognitive sciences, some of the most significant of which are human behavior simulation, structural simulation of the human brain, and designing interface systems between man and computer. Regarding the importance of cognitive disorders and problems in the society, the researches of the MLRG have been focused on early screening, autism rehabilitation, designing addiction simulation, Parkinson’s disease simulation, cognitive games, and simulating human decision-making. The result of many of these researches, such as early screening of autism by intelligent expert system, intelligent toy machine for early screening of autism, and intelligent games for screening of cognitive abilities have been made available to the public.
Pegah Soleyman, Bijan Mehralizadeh
Human Skeleton Detector
Children with autism often exhibit repetitive behaviors and activities such as arm flapping, head banging, or twirling an object over and over. Using Kinect, we have tried to recognize these behaviors by detecting the children’s skeletons. Ultimately, these movement patterns can be detected through algorithms of pattern recognition. What can be interesting for the audience is seeing their own skeletons as their make random movements. We intend to detect those movements and display them on a screen real-time.
Interactive games are among the tools that are used for the treatment of children with autism. Because of their simple design, these games have received considerable attention in the recent years. Children with autism have difficulty identifying and expressing their emotions. “Emotion Imitator” is a game that is designed to teach children about their emotions in different environments and with different games. The game includes an “avatar” section that detects the facial expression and imitates them.
Pegah Soleyman, Bijan Mehralizadeh, Shahab Nikkho, Ali Momen Sani
Interactive Smart Blue Parrot
Because robots can be used as toys while they can be playmates, companions, and even therapists, they have received much attention by researchers in the recent years. Robots can provide social assistance for autistic children, as well as other spheres of heath care such as being a companion for elderly people. “Interactive Smart Blue Parrot” has been designed to make verbal communication with autistic children. Parrots are known for their ability to imitate human speech, therefore they are attractive animals. Abilities like recognition of faces, objects, and speech can enhance the function of these robots as a substitute for humans in communication. Thus, these abilities have been included in this robot to some extent.
Smart Screening Car
Children with autism have repetitive behaviors. This interest in repetition can be seen in their games. One of main patterns of these repetitive behaviors is pulling, pushing, and rolling wheels over and over. If we can detect these repetitive actions with toy cars, we can use the data in the screening process of autistic children. Using the signals from an accelerometer installed inside a toy car, monitoring the output signals of autistic and normal children, and using pattern recognition algorithms, we have succeeded in detecting these children with relatively high precision. We can see the output signals while children play different games with the car.
Venue: Ava Center