Towards a Complete Set of Gym Exercises Detection Using Smartphone Sensors

Table 2

Detail of sensors.

Sensor name	Times used	Reason	Accuracies min., max., and avg.	References
Accelerometer	12	The accelerometer is the most powerful sensor in smartphones. It can be used for activity recognition by inferring the user’s movements, such as walking, standing, running, sitting, and gym activities.	Min. accuracy = 82.2%	[15, 23–33]
			Max. accuracy = 97.3%
			Avg. accuracy = 87.7%
Accelerometer and gyroscope	08	Accelerometer and a gyroscope, to be used in recognizing physical activity and providing the strongest result.	Min. accuracy = 67.8%	[10, 11, 31, 34–38]
			Max. accuracy = 97%
			Avg. accuracy = 88.3%
Accelerometer, gyroscope, and magnetometer	05	Adding a magnetometer with an accelerometer and gyroscope. The results are not encouraging because magnetometer causes overfitting in training classifiers due to its dependence on directions.	Min. accuracy = 71.6%	[10, 11, 34, 37, 39]
			Max. accuracy = 96%
			Avg. accuracy = 82.6%