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 an important family of methods to speed up deep model'sinference. Previous filter pruning algorithms regard channel pruning and modelfine-tuning as two independent steps. This paper argues that combining theminto a single end-to-end trainable system will lead to better results. Wepropose an efficient channel selection layer, namely AutoPruner, to find lessimportant filters automatically in a joint training manner. Our AutoPrunertakes previous activation responses as an input and generates a true binaryindex code for pruning. Hence, all the filters corresponding to zero indexvalues can be removed safely after training. We empirically demonstrate thatthe gradient information of this channel selection layer is also helpful forthe whole model training. By gradually erasing several weak filters, we canprevent an excessive drop in model accuracy. Compared with previousstate-of-the-art pruning algorithms (including training from scratch),AutoPruner achieves significantly better performance. Furthermore, ablationexperiments show that the proposed novel mini-batch pooling and binarizationoperations are vital for the success of filter pruning.

AutoPruner: An End-to-End Trainable Filter Pruning Method for Efficient Deep Model Inference