Human Pose Estimation and Action Recognition

Abstract:

This project explores computer vision applications with a focus on human pose estimation (HPE) and human action recognition (HAR). Unlike existing research that mainly addresses pure HPE, our study aims to bridge the gap by developing methodologies and datasets tailored for the required purpose. We will create a dedicated dataset, capturing unique interaction patterns and motion behaviors, to evaluate and advance algorithms in HPE and HAR under certain interesting environments.

Skills and experience required for the project:

Students are expected to have a background in computer science, artificial intelligence, statistics, data science, or engineering.
Programming skills in Python are required, with familiarity in machine learning and deep learning frameworks being highly desirable.
Experience in computer vision, data processing, and AI will be advantageous.
Soft skills such as teamwork, problem-solving, and attention to detail are also important.
Prior project work or research experience in machine learning or related fields will be considered a bonus.