General Objective: Develop new models for the recognition of 5 or more hand gestures using electromyographic (EMG) signals and artificial intelligence techniques with a high degree of classification and recognition accuracy and low computational cost in terms of processing time.
Specific objectives
- To analyze the statistical properties of the electromyographic (EMG) signals generated by the forearm muscles, as well as the mathematical models that describe this type of signals, through time series, to define the conditions and restrictions for their classification.
- Analyze the probabilistic and approximation properties of mathematical functions of artificial intelligence techniques applicable to the classification of time series to define a list of algorithms applicable to the classification of EMG signals.
- Define a protocol for the evaluation of the classification accuracy, recognition accuracy and the computational cost in time of gesture recognition models.
- Develop a specific or individual model, which is trainable by each user, for the recognition of 5 or more hand gestures using electromyographic (EMG) signals and artificial intelligence techniques.
- Develop a general model, usable by any user, for the recognition of 5 or more hand gestures using electromyographic (EMG) signals and artificial intelligence techniques.
- Implement in Matlab® the models developed for the recognition of the following gestures: hand to the left, hand to the right, fist, open hand, and double stroke of the fingers using electromyographic (EMG) signals acquired through a commercial sensor.
- Evaluate the classification accuracy, the recognition accuracy and the computational cost in time of the proposed models and make comparisons with each other and with models of similar purpose proposed in the scientific literature.
Participating Institutions:
EPN, ESPE, UTA.
Participants:
Director of the Benalcázar Palacios project Marco Enrique, PhD. In Electronic Engineering – EPN.
- Lorraine Isabel Barona Lopez
- Angel Leonardo Valdivieso Caraguay
- Jaramillo Yanez Andres Gabriel
- Rolando Marcelo Álvarez Veintimilla
- Freddy Geovanny Benalcazar Palacios
- Ruben Eduardo Nogales Goalkeeper
- Jaime Guilcapi Mosquera
Awarded budget: $63000
Project status: Signing of agreements.
