Human/ Animal Activity Recognition Data Analysis
Problem Statement:
Human or animal tracking and activity recognition can give us a bird-eye-view of their health and an early alarm if they are suffering from distress or anxiety. In this project, we will use available datasets of human/ animal activities collected by a millimeter-wave frequency modulated continuous wave radars (FMCW). The main goal of this project is to do a literature review about the most common Machin Learning and Deep Learning methods used in human/ animal activity recognition. If necessary, try to develop a neural network architecture that suits the best available data set. Finally, different methods should be compared based on accuracy and precision.
Task:
1. Literature review on the most common Machin Learning and Deep Learning methods used in human/ animal activity recognition.
2. Write Python scripts of the found methods, and design the most suitable neural network architecture. In the end all methods should be compared to each other.
Note: You are supposed to know how to write codes in Python language. Moreover, you should be acquainted with Machin Learning and Deep Learning methods.
Work:
%10 Theory, %70 Simulations, %20 Writing
Contact:
Ehsan Sam Sadeghi (E.sadeghi@utwente.nl)