Name
Abstract | ||
|---|---|---|
Image instance retrieval is the problem of retrieving images from a database which contain the same object. Convolutional Neural Network (CNN) based descriptors are becoming the dominant approach for generating global image descriptors for the instance retrieval problem. One major drawback of CNN-based global descriptors is that uncompressed deep neural network models require hundreds of megabytes of storage making them inconvenient to deploy in mobile applications or in custom hardware. In this work, we study the problem of neural network model compression focusing on the image instance retrieval task. We study quantization, coding, pruning and weight sharing techniques for reducing model size for the instance retrieval problem. We provide extensive experimental results on the trade-off between retrieval performance and model size for different types of networks on several data sets providing the most comprehensive study on this topic. We compress models to the order of a few MBs: two orders of magnitude smaller than the uncompressed models while achieving negligible loss in retrieval performance1. |
Year | DOI | Venue |
|---|---|---|
2017 | 10.1109/DCC.2017.93 | 2017 Data Compression Conference (DCC) |
Keywords | DocType | Volume |
CNN,Compression,Image Instance Retrieval | Conference | abs/1701.04923 |
ISSN | ISBN | Citations |
1068-0314 | 978-1-5090-6722-0 | 3 |
PageRank | References | Authors |
0.37 | 17 | 7 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Vijay Chandrasekhar | 1 | 191 | 22.83 |
| Jie Lin | 2 | 3495 | 502.80 |
| Qianli Liao | 3 | 44 | 7.08 |
| Olivier Morère | 4 | 62 | 4.56 |
| Antoine Veillard | 5 | 54 | 5.82 |
| Ling-yu Duan | 6 | 1770 | 124.87 |
| Tomaso Poggio | 7 | 13488 | 3380.01 |