The Mating Rituals of Deep Neural Networks: Learning Compact Feature Representations Through Sexual Evolutionary Synthesis

Audrey G. Chung, Mohammad Javad Shafiee, Paul Fieguth, Alexander Wong; Proceedings of the IEEE International Conference on Computer Vision (ICCV), 2017, pp. 1220-1227

Evolutionary deep intelligence was recently proposed as a method for achieving highly efficient deep neural network architectures over successive generations. Inspired by nature, we propose the incorporation of sexual evolutionary synthesis. Rather than the current asexual synthesis of networks, we aim to produce more compact feature representations by synthesizing more diverse and generalizable offspring networks in subsequent generations via the combination of two parent networks. Experimental results were obtained using the MNIST and CIFAR-10 datasets, and showed improved architectural efficiency and comparable testing accuracy relative to the baseline asexual evolutionary neural networks. In particular, the network synthesized via sexual evolutionary synthesis for MNIST had double the architectural efficiency (cluster efficiency of 34.29x and synaptic efficiency of 258.37x) in comparison to asexual evolutionary synthesis, with both networks achieving a testing accuracy of 97%.