Pruning Convolutional Neural Networks For Image Instance Retrieval

Abstract

In this work, we focus on the problem of image instance retrieval with deep descriptors extracted from pruned Convolutional Neural Networks (CNN). The objective is to heavily prune convolutional edges while maintaining retrieval performance. To this end, we introduce both data-independent and data-dependent heuristics to prune convolutional edges, and evaluate their performance across various compression rates with different deep descriptors over several benchmark datasets. Further, we present an end-to-end framework to fine-tune the pruned network, with a triplet loss function specially designed for the retrieval task. We show that the combination of heuristic pruning and fine-tuning offers 5x compression rate without considerable loss in retrieval performance.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keymanek2017pruning

Related papers

• Compression Of Deep Neural Networks For Image Instance Retrieval (2017)8.82
• Group Invariant Deep Representations For Image Instance Retrieval (2016)0.00
• Progressive Local Filter Pruning For Image Retrieval Acceleration (2020)12.74
• Fine-tuning CNN Image Retrieval With No Human Annotation (2017)22.75
• What Is The Best Practice For Cnns Applied To Visual Instance Retrieval? (2016)0.00
• From Selective Deep Convolutional Features To Compact Binary Representations For Image Retrieval (2018)10.35
• Faster R-CNN Features For Instance Search (2016)14.62
• Class-weighted Convolutional Features For Visual Instance Search (2017)12.81