Performing for the camera

Analyzing footage of yourself can be quite challenging. Some of us are overly critical of what they look like, and how they perform. These people tend to underestimate their own achievements and performances. Others are completely ignorant/oblivious to how they look and perform, causing them overestimate how well they did whatever they were doing. This can especially be a problem when trying to learn a physical skill such as skiing.

Visma ConsultIoTing has deviced a solution that helps the athlete be objective when analyzing their own performance. By using computer vision and machine learning to perform pose estimation of camera footage, athletes can objectively judge their technique and performance. Answering questions like "Is the angle of my back right?", "Am I leaning far enough forward?", and "How does my pose look?" can be tricky, especially if you have a distorted view of yourself.

Pose recognition gives an objective view of the athletes performance, and connecting the different joints by straight lines simplifies a lot of the challenges when it comes to evaluating skiing technique. This solution enables the user both to get real time feedback, as well as remote recording and doing post-workout analysis. An example of the former can be seen in the following footage.

Skiing with smart vision

To see an example of the latter, please stay tuned for our interactive dashboard!

This post is a submission to receive the "Eyes wide shutter"-badge