Strengthening muscles can reduce body fat, increase lean muscle mass, maintain independence while aging, manage chronic conditions, and improve balance, reducing the risk of falling. The most critical factor inducing effectiveness in strength training is neuromuscular connection by adopting attentional focus during training. However, this is troublesome for end users since numerous fitness tracking devices or applications do not provide the ability to track the effectiveness of users' workout at the neuromuscular level. A practical approach for detecting attentional focus by assessing neuromuscular activity through biosignals has not been adequately evaluated. The challenging task to make the idea work in a real-world scenario is to minimize the cost and size of the clinical device and use a recognition system for muscle contraction to ensure a good user experience. We then introduce a multitasking and multiclassification network and an EMG shirt attached with noninvasive sensing electrodes that firmly fit to the body's surface, measuring neuron muscle activity during exercise. Our study exposes subjects to standard free-weight exercises focusing on isolated and compound muscle on the upper limb. The results of the experiment show a 94.79% average precision at different maximum forces of attentional focus conditions. Furthermore, the proposed system can perform at different lifting weights of 67% and 85% of a person's 1RM to recognize individual exercise effectiveness at the muscular level, proving that adopting attentional focus with low-intensity exercise can activate more upper-limb muscle contraction.
Keywords: attentional focus; biosignal sensing; exercise monitoring; neuron network; wearable device.