English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Channel pruning is a popular technique for compressing convolutional neuralnetworks (CNNs), where various pruning criteria have been proposed to removethe redundant filters. From our comprehensive experiments, we found two blindspots in the study of pruning criteria: (1) Similarity: There are some strongsimilarities among several primary pruning criteria that are widely cited andcompared. According to these criteria, the ranks of filters'Importance Scoreare almost identical, resulting in similar pruned structures. (2)Applicability: The filters'Importance Score measured by some pruning criteriaare too close to distinguish the network redundancy well. In this paper, weanalyze these two blind spots on different types of pruning criteria withlayer-wise pruning or global pruning. The analyses are based on the empiricalexperiments and our assumption (Convolutional Weight Distribution Assumption)that the well-trained convolutional filters each layer approximately follow aGaussian-alike distribution. This assumption has been verified throughsystematic and extensive statistical tests.

Convolution-Weight-Distribution Assumption: Rethinking the Criteria of Channel Pruning