Search our site
Search our site

Mr Fariborz Baghaei Naeini

Research project: Human Action Recognition using Event-Driven Cameras

Abstract

Recognition techniques in computer vision have a broad range of applications to interpret human activities, gestures, identification, emotions, and posture recognition. This study will mainly focus on human actions such as walking, jogging, waving hands, turning around, kicking, punching and composite actions. Many approaches have been using additional equipment such as gloves, stylus, and other position trackers to build and improve recognition applications. However, development of cameras and computer vision algorithms offer this opportunity to detect, track, and recognise human activities accurately without additional equipment. Dynamic Active-pixel Vision Sensor (DAVIS) is a hybrid sensor, that in addition to a standard frame-based camera, it can also work as an event-based, low-latency and low-power sensor that can be used to achieve solutions for human activity recognition where latency power consumption may be important. This project will investigate novel activity recognition techniques that will exploit the DAVIS camera and will investigate the optimal compromise between latency, speed, power consumption and recognition accuracy.

Biography

Dedicated and highly self-motivated Ph.D. candidate in Artifical Intelligence with hands-on experience in image processing, machine learning, and robotics.  In addition to my PhD research, I am working on robotic grasping systems to detect incipient slip and vibration using vision-based techniques.  Moreover, I worked as a collaborative member of digital signal processing team to study Figures of Merits of a 3-D Microwave Imaging System (MARIA) for breast cancer diagnosis for my MSc dissertation. I am so passionate about researching in multi-disciplinary topics  as a team and applying vision-based and artificial intelligence methods in various fields such as chemistry reactions analysis, autonomous cars applications, and high speed vision-based control systems. So, please do not hesitate and contact me if you are interested in collaborating in these research fields.

Areas of research interest

  • Human Action Recognition
  • Gesture Recognition
  • Machine Learning
  • Computer Vision
  • Robotics

Qualifications

  • Msc. in Embeded Systems with Management Studies, Kingston University London
  • BSc. in Computer Hardware Engineering, Azad University of Tehran Central Branch
Site menu