keyword
fully convolutional Siamese networks
A fully convolutional Siamese network is a deep learning architecture that combines twin weight-sharing neural network branches with a fully convolutional design to compare two input images or patches. Because the network consists entirely of convolutional operations without fully connected layers, it exhibits translation equivariance and can accept inputs of arbitrary or differing spatial dimensions. The parallel branches extract high-level feature representations from each input, which are then combined using operations such as cross-correlation, concatenation, or feature subtraction to generate a dense spatial response map. This design allows for efficient, end-to-end similarity evaluation across entire image areas in a single forward pass, making it widely applied in computer vision tasks including visual object tracking, change detection, and template matching.
2 items
Fully Convolutional Siamese Networks for Change Detection
Rodrigo Caye Daudt, Bertrand Le Saux, Alexandre Boulch
Why you should read this
Proposes fully convolutional Siamese architectures for remote sensing change detection that train from scratch on coregistered image pairs to deliver superior accuracy at over 500 times the speed of previous methods.
This paper presents three fully convolutional neural network architectures which perform change detection using a pair of coregistered images. Most notably, we propose two Siamese extensions of fully convolutional networks which use heuristics about the current problem to achieve the best results in our tests on two open change detection datasets, using both RGB and multispectral images. We show that our system is able to learn from scratch using annotated change detection images. Our architectures achieve better performance than previously proposed methods, while being at least 500 times faster than related systems. This work is a step towards efficient processing of data from large scale Earth observation systems such as Copernicus or Landsat.
Added
2026-09-24
Fully-Convolutional Siamese Networks for Object Tracking
Luca Bertinetto, Jack Valmadre, João F. Henriques, Andrea Vedaldi, Philip H. S. Torr
Why you should read this
Introduces a fully-convolutional Siamese network for visual object tracking that replaces slow online model adaptation with end-to-end offline learning, delivering state-of-the-art tracking accuracy at speeds well beyond real time.
The problem of arbitrary object tracking has traditionally been tackled by learning a model of the object's appearance exclusively online, using as sole training data the video itself. Despite the success of these methods, their online-only approach inherently limits the richness of the model they can learn. Recently, several attempts have been made to exploit the expressive power of deep convolutional networks. However, when the object to track is not known beforehand, it is necessary to perform Stochastic Gradient Descent online to adapt the weights of the network, severely compromising the speed of the system. In this paper we equip a basic tracking algorithm with a novel fully-convolutional Siamese network trained end-to-end on the ILSVRC15 dataset for object detection in video. Our tracker operates at frame-rates beyond real-time and, despite its extreme simplicity, achieves state-of-the-art performance in multiple benchmarks.
Added
2026-09-16