The development of sensors for human-machine interfaces is essential for various applications in robotics, assistive technologies, rehabilitation, and medicine. Gestures recognition has great importance for monitoring and integration with actuators; however, most commercial sensors are expensive and complex, which limits their operation by non-high skilled users. In this context, the force myography technique (FMG), which characterizes the stimuli of forearm muscles in terms of mechanical pressures, emerges as an alternative to the surface electromyography. In this work, an optical fiber sensor based on the FMG technique was proposed for identifying the hand gestures, wherein the interrogation system was integrated into a smartphone to provide a simplified and intuitive user interface. The smartphone flashlight excites a pair of polymer optical fibers coupled with a 3D-printed case, whereas the output signal is detected by the camera. The light intensity is modulated through wearable force-driven microbending transducers placed in the forearm. Subsequently, the acquired optical signals are processed by an algorithm based on decision trees and residual error. The classifier was tested for different thresholds from 3 to 20% of the light intensity signal, and the best results were verified for 8% cutoff limit. Furthermore, the receiver operating characteristic (ROC) curves were plotted for each pattern and generated areas under the ROC curves near the unit. The sensor provided a hit rate of 87% regarding four postures, yielding reliable performance with a simple, portable, and low-cost setup embedded on a smartphone.