CVFeb 26, 2022
Symmetric Convolutional Filters: A Novel Way to Constrain Parameters in CNN
arXiv:2202.13099v1
Originality Incremental advance
AI Analysis
This addresses parameter efficiency for CNN users, but appears incremental as it builds on existing pruning techniques.
The paper tackles the problem of parameter redundancy in CNNs by introducing symmetric convolutional filters, resulting in effective generalization and structured parameter elimination.
We propose a novel technique to constrain parameters in CNN based on symmetric filters. We investigate the impact on SOTA networks when varying the combinations of symmetricity. We demonstrate that our models offer effective generalisation and a structured elimination of redundancy in parameters. We conclude by comparing our method with other pruning techniques.