Convolutional Neural Network Compression through Generalized Kronecker Product Decomposition

Authors: Marawan Gamal Abdel Hameed, Marzieh S. Tahaei, Ali Mosleh, Vahid Partovi Nia771-779

AAAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results for image classification on CIFAR-10 and Image Net datasets using Res Net, Mobile Netv2 and Se Net architectures substantiate the effectiveness of our proposed approach. We find that GKPD outperforms state-of-the-art decomposition methods including Tensor-Train and Tensor-Ring as well as other relevant compression methods such as pruning and knowledge distillation.
Researcher Affiliation Collaboration Marawan Gamal Abdel Hameed 1,2*, Marzieh S. Tahaei1 , Ali Mosleh1, Vahid Partovi Nia1 1Noah s Ark Lab, Huawei Technologies Canada 2University of Waterloo
Pseudocode Yes Algorithm 1: Forward Pass
Open Source Code No The paper does not provide an explicit statement or link for open-source code for the described methodology.
Open Datasets Yes Experimental results for image classification on CIFAR-10 and Image Net datasets... Table 1 shows the top-1 accuracy on the CIFAR-10 (Krizhevsky 2009) dataset... on a larger scale dataset i.e, Image Net (Krizhevsky, Sutskever, and Hinton 2012).
Dataset Splits Yes Experimental results for image classification on CIFAR-10 and Image Net datasets... Top-1 accuracy measured on CIFAR-10... Top-1 accuracy measured on Image Net. These are well-established benchmark datasets with standard train/validation/test splits, which are implicitly used for evaluating performance.
Hardware Specification No The paper does not explicitly describe the specific hardware (e.g., GPU models, CPU types) used for running its experiments.
Software Dependencies No The paper does not provide specific software dependencies with version numbers.
Experiment Setup No The paper states 'We provide implementation details in the Supplementary Material' but does not include specific hyperparameters or training settings in the main text.