Leveraging Filter Correlations for Deep Model Compression

We present a filter correlation based model compression approach for deep convolutional neural networks. Our approach iteratively identifies pairs of filters with the largest pairwise correlations and drops one of the filters from each such pair. However, instead of discarding one of the filters fro...

Full description

Saved in:
Bibliographic Details
Main Authors: Singh, Pravendra, Verma, Vinay Kumar, Rai, Piyush, Namboodiri, Vinay P.
Format: Conference Proceeding
Language:English
Subjects:
Acceleration
Correlation
Memory management
Quantization (signal)
Redundancy
Task analysis
Transmission line measurements
Citations: Items that cite this one
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We present a filter correlation based model compression approach for deep convolutional neural networks. Our approach iteratively identifies pairs of filters with the largest pairwise correlations and drops one of the filters from each such pair. However, instead of discarding one of the filters from each such pair naïvely, the model is re-optimized to make the filters in these pairs maximally correlated, so that discarding one of the filters from the pair results in minimal information loss. Moreover, after discarding the filters in each round, we further finetune the model to recover from the potential small loss incurred by the compression. We evaluate our proposed approach using a comprehensive set of experiments and ablation studies. Our compression method yields state-of-the-art FLOPs compression rates on various benchmarks, such as LeNet-5, VGG-16, and ResNet-50,56, while still achieving excellent predictive performance for tasks such as object detection on benchmark datasets.
ISSN:2642-9381
DOI:10.1109/WACV45572.2020.9093331