Optimizing Android Facial Expressions Using Genetic Algorithms

Abstract

Because the internal structure, degree of freedom, skin control position and range of the android face are different, it is very difficult to generate facial expressions by applying existing facial expression generation methods. In addition, facial expressions differ among robots because they are designed subjectively. To address these problems, we developed a system that can automatically generate robot facial expressions by combining an android, a recognizer capable of classifying facial expressions and a genetic algorithm. We have developed two types (older men and young women) of android face robots that can simulate human skin movements. We selected 16 control positions to generate the facial expressions of these robots. The expressions were generated by combining the displacements of 16 motors. A chromosome comprising 16 genes (motor displacements) was generated by applying real-coded genetic algorithms; subsequently, it was used to generate robot facial expressions. To determine the fitness of the generated facial expressions, expression intensity was evaluated through a facial expression recognizer. The proposed system was used to generate six facial expressions (angry, disgust, fear, happy, sad, surprised); the results confirmed that they were more appropriate than manually generated facial expressions.

Smile Like You Mean It: Driving Animatronic Robotic Face with Learned Models

“…Hyung et al [18] have proposed a genetic algorithms to search for the best facial expression. However, the algorithm is inefficient when the search space grows with the complexity of the robot.…”

“…Some traditional methods [10]- [16] define a set of pre-specified facial expressions. Others generalize this process to search for closest match from a database [17] or by following an fitness function [17], [18]. However, as human expressions are highly diverse, these approaches have limited value in practical robot-human interactions.…”

Optimizing Facial Expressions of an Android Robot Effectively: a Bayesian Optimization Approach

“…Ren et al [14] proposed a kinematics-based learning method for robot XIN-REN to imitate expression of a human subject. Hyung et al [13] used a genetic algorithm to identify optimal facial expressions in the search space. Similarly, Habib [20] proposed a learningbased method with a genetic algorithm capable of inverse nonlinear mapping from human faces to robot actuators.…”

Facial Expression Recognition with Fusion Features Extracted from Salient Facial Areas

“…(1) replicating the intricate biomechanics of human facial musculature [7][8][9][10][11][12][13][14], and (2) generating nuanced human expressions through responsive algorithms based on advanced imitation learning [15][16][17][18][19]. Overcoming these challenges necessitates a comprehensive approach that integrates embodiment design and motion synthesis.…”

“…The approach comprises three major components: (1) skinning-oriented robot development designs and constructs the animatronic face paired with a kinematics simulator based on the target skinning appearance, (2) skinning motion imitation learning involves learning an LBS-based model from 3D human demonstrations to generate facial motions from speech, and (3) speech-driven robot orchestration generates animatronic facial expressions during inference by utilizing the developed platform, simulator, and learned model. algorithms [16], visual mimicry learning [17], Bayesian optimization [18], and MAP-Elites algorithms [19]. While these approaches have greatly enhanced the flexibility of animatronic robot faces, they remain fundamentally constrained by their reliance on mimicking observed human expressions through a master-slave mapping paradigm.…”