%0 Journal Article %D 2017 %T Compression of Deep Neural Networks for Image Instance Retrieval %A Vijay Chandrasekhar %A Jie Lin %A Qianli Liao %A Olivier Morère %A Antoine Veillard %A Lingyu Duan %A Tomaso Poggio %X

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 {\it global image descriptors} for the instance retrieval problem. One major drawback of CNN-based {\it 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 performance.

%8 01/2017 %G eng %U https://arxiv.org/abs/1701.04923 %0 Journal Article %D 2017 %T Pruning Convolutional Neural Networks for Image Instance Retrieval %A Gaurav Manek %A Jie Lin %A Vijay Chandrasekhar %A Lingyu Duan %A Sateesh Giduthuri %A Xiaoli Li %A Tomaso Poggio %K CNN %K Image Instance Re- trieval %K Pooling %K Pruning %K Triplet Loss %X

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.

%8 07/2017 %G eng %U https://arxiv.org/abs/1707.05455 %0 Generic %D 2016 %T Group Invariant Deep Representations for Image Instance Retrieval %A Olivier Morère %A Antoine Veillard %A Jie Lin %A Julie Petta %A Vijay Chandrasekhar %A Tomaso Poggio %X

Most image instance retrieval pipelines are based on comparison of vectors known as global image descriptors between a query image and the database images. Due to their success in large scale image classification, representations extracted from Convolutional Neural Networks (CNN) are quickly gaining ground on Fisher Vectors (FVs) as state-of-the-art global descriptors for image instance retrieval. While CNN-based descriptors are generally remarked for good retrieval performance at lower bitrates, they nevertheless present a number of drawbacks including the lack of robustness to common object transformations such as rotations compared with their interest point based FV counterparts.


In this paper, we propose a method for computing invariant global descriptors from CNNs. Our method implements a recently proposed mathematical theory for invariance in a sensory cortex modeled as a feedforward neural network. The resulting global descriptors can be made invariant to multiple arbitrary transformation groups while retaining good discriminativeness.


Based on a thorough empirical evaluation using several publicly available datasets, we show that our method is able to significantly and consistently improve retrieval results every time a new type of invariance is incorporated. We also show that our method which has few parameters is not prone to over fitting: improvements generalize well across datasets with different properties with regard to invariances. Finally, we show that our descriptors are able to compare favourably to other state-of-theart compact descriptors in similar bitranges, exceeding the highest retrieval results reported in the literature on some datasets. A dedicated dimensionality reduction step –quantization or hashing– may be able to further improve the competitiveness of the descriptors.

%8 01/2016 %G English %1

arXiv:1601.02093v1

%2

http://hdl.handle.net/1721.1/100796

%0 Journal Article %J arXiv.org %D 2016 %T Nested Invariance Pooling and RBM Hashing for Image Instance Retrieval %A Olivier Morère %A Antoine Veillard %A Vijay Chandrasekhar %A Tomaso Poggio %K CNN %K Hashing %K Image Instance Retrieval %K Invariant Representation %K Regularization %K unsupervised learning %X

The goal of this work is the computation of very compact binary hashes for image instance retrieval. Our approach has two novel contributions. The first one is Nested Invariance Pooling (NIP), a method inspired from i-theory, a mathematical theory for computing group invariant transformations with feed-forward neural networks. NIP is able to produce compact and well-performing descriptors with visual representations extracted from convolutional neural networks. We specifically incorporate scale, translation and rotation invariances but the scheme can be extended to any arbitrary sets of transformations. We also show that using moments of increasing order throughout nesting is important. The NIP descriptors are then hashed to the target code size (32-256 bits) with a Restricted Boltzmann Machine with a novel batch-level regularization scheme specifically designed for the purpose of hashing (RBMH). A thorough empirical evaluation with state-of-the-art shows that the results obtained both with the NIP descriptors and the NIP+RBMH hashes are consistently outstanding across a wide range of datasets.

%B arXiv.org %8 03/2016 %G eng %U https://arxiv.org/abs/1603.04595