Towards Lightweight Neural Animation : Exploration of Neural Network Pruning in Mixture of Experts-based Animation Models

TL;DR

This paper explores neural network pruning techniques to compress Mixture of Experts-based neural animation models, reducing computational costs while maintaining motion quality and minimizing artifacts.

Contribution

It introduces pruning algorithms applied to MLP-MoE models for character animation, balancing efficiency and motion fidelity in real-time applications.

Findings

Pruned models have fewer parameters and faster computation.

Motion quality remains high with fewer artifacts after pruning.

High-level motion features are preserved despite pruning.

Abstract

In the past few years, neural character animation has emerged and offered an automatic method for animating virtual characters. Their motion is synthesized by a neural network. Controlling this movement in real time with a user-defined control signal is also an important task in video games for example. Solutions based on fully-connected layers (MLPs) and Mixture-of-Experts (MoE) have given impressive results in generating and controlling various movements with close-range interactions between the environment and the virtual character. However, a major shortcoming of fully-connected layers is their computational and memory cost which may lead to sub-optimized solution. In this work, we apply pruning algorithms to compress an MLP- MoE neural network in the context of interactive character animation, which reduces its number of parameters and accelerates its computation time with a trade-off between this acceleration and the synthesized motion quality. This work demonstrates that, with the same number of experts and parameters, the pruned model produces less motion artifacts than the dense model and the learned high-level motion features are similar for both