High performance inference of gait recognition models on embedded systems

Edge computing is gaining importance in the realm of Deep Learning, particularly after powerful devices such as recent heterogeneous embedded systems have demonstrated remarkable skills for accelerating their chal-lenging computational requirements. In this work, we evaluate different hardware and software optimizations applied to state-of-the-art gait recognition approaches deployed on two Jetson devices with very different hardware capabilities. Specifically, we have selected three models with different characteristics in order to provide an in-depth deployment evaluation. This way, a 2D convolution-based model allows us to evaluate devices performance when a huge number of parameters must be managed. A model based on 3D convolutions is deployed to study devices capability to perform these kinds of operations. Finally, a novel model with a small number of parameters but with a huge number of activations is also evaluated. Obtained results show that different hardware and software optimizations are able to improve up to 5.4x energy consumption and 4.2x execution time w.r.t. baseline deployment, depending on the model and target device.1