Loading...
- Type of Document: M.Sc. Thesis
- Language: Farsi
- Document No: 48816 (05)
- University: Sharif University of Technology
- Department: Electrical Engineering
- Advisor(s): Fatemizadeh, Emad
- Abstract:
- The definition of human action recognition is classification of input visual elements based on the action which is done by a person in the scene. One of the most important topics in the filed which has lots of applications is human action recognition in videos. Some of these applications are surveillance, video retrieval, human computer interaction and smart houses. Due to increments in number of alone elderly people, surveillance of them is one of the important applications of human action recognition. The challenges of the task are, camera movement, differences of environment and differences in acting by different actors.The goal of the project is proposing a deep convolutional neural network based algorithm to recognize different human actions better than previous works on the well known dataset in the field and finally proposing a fully automatic system to work in smart houses, with the goal of recognizing elderly people’s actions. the most important feature of the system is the ability of working in different environments without need to any calibration and tuning. To do so we use convolutional neural networks due to their most important abilities such as, being proposed based on the human visual system, their promising performance on recognition tasks, having end-to-end training and performing real time after training is done. Achieving an accuracy of 84.9 % and the speed of 237 frames per seconds by the proposed network on the UCF101 action dataset verify the claim
- Keywords:
- Human Action Recognition ; Deep Learning ; Smart Home ; Video Data Classification ; Convolutional Neural Network
- محتواي کتاب
- view
- مقدمه
- مرور ادبیات
- روش پیشنهادی
- نتایج و بحث
- نتیجهگیری و پیشنهادها
- پیوستها
- جزئیات لایههای شبکه عصبی پيچشی