Abstract:A number of recent studies have shown that a Deep Convolutional Neural Network (DCNN) pretrained on a large dataset can be adopted as a universal image description which leads to astounding performance in many visual classification tasks. Most of these studies, if not all, adopt activations of the fully-connected layer of a DCNN as the image or region representation and it is believed that convolutional layer activations are less discriminative. This paper, however, advocates that if used appropriately convolutional layer activations can be turned into a powerful image representation which enjoys many advantages over fully-connected layer activations. This is achieved by adopting a new technique proposed in this paper called cross-convolutional-layer pooling. More specifically, it extracts subarrays of feature maps of one convolutional layer as local features and pools the extracted features with the guidance of feature maps of the successive convolutional layer. Compared with exising methods that apply DCNNs in the local feature setting, the proposed method is significantly faster since it requires much fewer times of DCNN forward computation. Moreover, it avoids the domain mismatch issue which is usually encountered when applying fully connected layer activations to describe local regions. By applying our method to four popular visual classification tasks, it is demonstrated that the proposed method can achieve comparable or in some cases significantly better performance than existing fully-connected layer based image representations while incurring much lower computational cost.

Delving into Fully Convolutional Networks Activations for Visual Recognition.

Aggregating Hierarchical Binary Activations for Image Retrieval

Exploiting Hierarchical Activations of Neural Network for Image Retrieval.

On the Exploration of Convolutional Fusion Networks for Visual Recognition.

Cross-convolutional-layer Pooling for Generic Visual Recognition.

The Treasure beneath Convolutional Layers: Cross-convolutional-layer Pooling for Image Classification

The Treasure Beneath Convolutional Layers: Cross-Convolutional-Iayer Pooling For Image Classification

Fusion that matters: convolutional fusion networks for visual recognition

Multiactivation Pooling Method in Convolutional Neural Networks for Image Recognition.

Bag of Surrogate Parts Feature for Visual Recognition

Dense Attention Convolutional Network for Image Classification

Fully Convolutional Networks for Action Recognition

Convolutional Neural Network with Nonlinear Competitive Units.

UP-CNN: Un-pooling Augmented Convolutional Neural Network.

Convolutional Networks with Cross-Layer Neurons for Image Recognition

Fully Convolutional Attention Networks for Fine-Grained Recognition

On the Large-Scale Transferability of Convolutional Neural Networks.

Cross-Level: A Practical Strategy for Convolutional Neural Networks Based Image Classification

Self-Adaption Multi-classifier Fusion Networks for Image Recognition

Multilevel Features Fusion In Deep Convolutional Neural Networks

Cross-Modal Retrieval with CNN Visual Features: A New Baseline