CupNet -- Pruning a network for geometric data
arXiv:2005.05276v2
Originality Synthesis-oriented
AI Analysis
This work addresses network pruning for geometric data, but it appears incremental as it applies an existing pruning approach to a specific domain without broad implications.
The paper tackled the problem of pruning neural networks by leveraging the geometric structure of cup meshes from simulated drawing data, achieving effective pruning in a straightforward manner.
Using data from a simulated cup drawing process, we demonstrate how the inherent geometrical structure of cup meshes can be used to effectively prune an artificial neural network in a straightforward way.