#PoseTrainer
In the modern world of recluse and recline, physical activity and improvement have taken the back burner. One of the critical components of such a lifestyle is the lack of a companion and a guide in the same. Based on the advent of Machine Learning and the vastly increasing online resources for physical activities, we propose PoseShaper. A companion tool for the recognition, identification and improvement of physical fitness and posture even amidst our daily lives. The model can easily be extended and modified to incorporate several more physical features and activities at any time to encompass the ever-growing possibilities of a better mind and body.
Human Pose Estimation(HPE) is the method of identifying and classifying human joints(key points) on the human body and correlating the same with physical postures and activities. The Estimation model intends to capture a set of key points on the body and classify the said points with respect to each other to establish a pair. A collection of said pairs can then further be used to calculate the respective vectors and thereby enable classification. The connections formed must be significant and thus not all possible key-points can be considered to correlate to form a connection.