Multilevel Context Representation for Improving Object Recognition
This work addresses the problem of enhancing object recognition accuracy for computer vision applications, but it is incremental as it builds upon existing architectures like AlexNet and GoogLeNet.
The paper tackled the problem of improving object recognition in CNNs by proposing a method that combines low- and high-level context, specifically extending AlexNet and GoogLeNet with additional connections in top layers. The result was a 1-2% relative reduction in classification error on ImageNet without increasing computational cost, and it was shown to be orthogonal to test data augmentation techniques, reducing runtime by 144 times during testing.
In this work, we propose the combined usage of low- and high-level blocks of convolutional neural networks (CNNs) for improving object recognition. While recent research focused on either propagating the context from all layers, e.g. ResNet, (including the very low-level layers) or having multiple loss layers (e.g. GoogLeNet), the importance of the features close to the higher layers is ignored. This paper postulates that the use of context closer to the high-level layers provides the scale and translation invariance and works better than using the top layer only. In particular, we extend AlexNet and GoogLeNet by additional connections in the top $n$ layers. In order to demonstrate the effectiveness of the proposed approach, we evaluated it on the standard ImageNet task. The relative reduction of the classification error is around 1-2% without affecting the computational cost. Furthermore, we show that this approach is orthogonal to typical test data augmentation techniques, as recently introduced by Szegedy et al. (leading to a runtime reduction of 144 during test time).