The researchers at Carnegie Mellon University have discovered that a stationary camera could be better at monitoring gym exercises than wearables.
More people now depend on smartwatches as a motivational tool for fitness. According to recent data, the sales of smartwatches has increased by 51 percent within the last 12 months. Also, 16 percent of adults in the United States now own a smartwatch.
But, as useful the current wearables are at cardio exercise, they do not sense all body motion effectively. The sensors are still not advanced and reliable enough to recognize the type of training and count repetitions.
Also, smartwatches depend entirely on what part of the body it’s worn. So, while it may be useful for dumbbell lifts, the device is completely useless of leg presses.
That’s where the GymCam comes in. The vision-based system developed by Ph.D. students in CMU’s Human-Computer Interaction Institute (HCII) can easily detect exercises in the gym.
In a statement, assistant professor in the HCII and Institute for Software Research, Mayank Goel said:
“If you are moving both your arms, you tend to move them together in time. However, if two people are exercising next to each other and performing the same exercise, they are usually not in sync, and we can tell the difference between them.”
Using the GymCam to Monitor Gym Exercises

GymCam requires only motion information. That means, the camera feed can be reduced to pixel-by-pixel changes protect the privacy of the person working out.
Also, the researchers suggested that reliance on motion information addresses the downside of using a single camera in a crowded gym – the inability to see a person’s whole body.
In a gym with a single camera, equipment, and other people can obstruct the camera’s view. However, GymCam only has to see a repetitive movement of any part of the exerciser’s body to detect the exercise.
Also, the system can detect a gym equipment location and use the individual’s position and movement to track what they’re doing.
Although the researchers developed the algorithm to work in a crowded gym, they suggested that it can work on a smartphone as well. That means users can track and record their workouts at home too.
But, the system’s function could extend beyond the physical exercise.
When combined with smartwatches, GymCam could help people with visual disabilities, navigate public spaces such as airports, shopping malls, etc. Instead of identifying the person’s face, the camera would track people using their motion signature
What’s more, people can choose to opt-out of being tracked or located, says Goel.
The researchers presented their findings at the International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp 2019) in London.
Comments (0)
Most Recent