Fully Convolutional Network and Region Proposal for Instance Identification With Egocentric Vision

Maxime Portaz, Matthias Kohl, Georges Quénot, Jean-Pierre Chevallet; Proceedings of the IEEE International Conference on Computer Vision (ICCV), 2017, pp. 2383-2391

This paper presents a novel approach for egocentric image retrieval and object detection. This approach uses fully convolutional network (FCN) to obtain region proposals without the need for an additional component in the network and training. It is particularly suited for small dataset with low object variability. The proposed network can be trained end-to-end and it produces an effective global descriptor as image representation. Additionally, it can be built upon any type of CNN pre-trained for classification. Through multiple experiments on two egocentric images datasets taken from museum visits, we show that the descriptor obtained using our proposed network outperforms those from previous state-of-the-art approaches. It is also just as memoryefficient, making it adapted to mobile device like augmented museum audio-guide.