Object Detectors Emerge in Deep Scene CNNs
This work provides insights into how CNNs learn hierarchical features, potentially benefiting computer vision researchers by revealing emergent object detection capabilities from scene classification tasks.
The authors tackled the problem of understanding representations learned by inner layers of deep convolutional neural networks (CNNs) trained for scene classification, and found that object detectors emerge automatically as a result, enabling the same network to perform both scene recognition and object localization in a single forward-pass without explicit object training.
With the success of new computational architectures for visual processing, such as convolutional neural networks (CNN) and access to image databases with millions of labeled examples (e.g., ImageNet, Places), the state of the art in computer vision is advancing rapidly. One important factor for continued progress is to understand the representations that are learned by the inner layers of these deep architectures. Here we show that object detectors emerge from training CNNs to perform scene classification. As scenes are composed of objects, the CNN for scene classification automatically discovers meaningful objects detectors, representative of the learned scene categories. With object detectors emerging as a result of learning to recognize scenes, our work demonstrates that the same network can perform both scene recognition and object localization in a single forward-pass, without ever having been explicitly taught the notion of objects.