A Low Effort Approach to Structured CNN Design Using PCA
share
Deep learning models hold state of the art performance in many fields, yet their design is still based on heuristics or grid search methods. This work proposes a method to analyze a trained network and deduce a redundancy-free, compressed architecture that preserves accuracy while keeping computational costs tractable. Model compression is an active field of research that targets the problem of realizing deep learning models in hardware. However, most pruning methodologies tend to be experimental, requiring large compute and time intensive iterations. We introduce structure into model design by proposing a single shot analysis of a trained network that frames model compression as a dimensionality reduction problem. The analysis outputs an optimal number of layers, and number of filters per layer without any iterative retraining procedures. We demonstrate our methodology on AlexNet and VGG style networks on the CIFAR-10, CIFAR-100 and ImageNet datasets.
READ FULL TEXT