Artificial Intelligence

Collective intelligence in biological groups can be employed to inspire the control of artificial complex systems, such as swarm robotics. However, modeling for the social interactions between individuals is still a challenging task. Without loss of generality, we propose a deep attention network model that incorporates the principles of biological Hard Attention mechanisms, that means an individual only pay attention to one or two neighbors for collective motion decision in large group.

Collective intelligence in biological groups can be employed to inspire the control of artificial complex systems, such as swarm robotics. However, modeling for the social interactions between individuals is still a challenging task. Without loss of generality, we propose a deep attention network model that incorporates the principles of biological Hard Attention mechanisms, that means an individual only pay attention to one or two neighbors for collective motion decision in large group.

Collective intelligence in biological groups can be employed to inspire the control of artificial complex systems, such as swarm robotics. However, modeling for the social interactions between individuals is still a challenging task. Without loss of generality, we propose a deep attention network model that incorporates the principles of biological Hard Attention mechanisms, that means an individual only pay attention to one or two neighbors for collective motion decision in large group.

Contains the benchmark Bayesian network dataset, which uses the seed of Bayesian networks from https://www.bnlearn.com. Some of the data comes from https://pages.mtu.edu/~lebrown/supplements/mmhc_paper/mmhc_index.html. And other datasets from the UCI that contain mixed data. These data can be used to learn the basic structure of Bayesian networks, the research of cause-based feature selection algorithms, etc bnlearn is an R package for learning the graphical structure of Bayesian networks, estimating their parameters and performing some useful inference.

Fundus Image Myopia Development (FIMD) dataset contains 70 retinal image pairs, in which, there is obvious myopia development between each pair of images. In addition, each pair of retinal images has a large overlap area, and there is no other retinopathy. In order to perform a reliable quantitative evaluation of registration results, we follow the annotation method of Fundus Image Registration (FIRE) dataset [1] to label control points between the pair of retinal images with the help of experienced ophthalmologists. Each image p

This LTE_RFFI project sets up an LTE device radio frequency fingerprint identification system using deep learning techniques. The LTE uplink signals are collected from ten different LTE devices using a USRP N210 in different locations. The sampling rate of the USRP is 25 MHz. The received signal is resampled to 30.72 MHz in Matlab and is saved in the MAT file form. The corresponding processed signals are included in the dataset. More details about the datasets can be found in the README document.

The Marketable Foods (MF) dataset was originally constructed to fine-tune the language and visual network layers and facilitates backdoor injections in text-to-image generative models. The dataset consists of images from three popular food corporations with prominent, recognisable brands (Coffee = Starbucks, Burger = McDonald's, Drink = Coca Cola). Samples were collected from the internet and were cleaned using a filtering algorithm discussed in the corresponding paper.

Weather radar echo extrapolation is an important approach for convective nowcasting, which predicts the evolution of convective systems in a short term. In recent years, radar echo extrapolation approaches based on deep learning have made significant progress and have been widely applied for radar echo extrapolation.

RITA (Resource for Italian Tests Assessment), is a new NLP dataset of academic exam texts written in Italian by second-language learners for obtaining the CEFR certification of proficiency level.
RITA dataset is available for automatic processing in CSV and XML format, under an agreement of citation.

To thoroughly investigate the non-overlapping registration problem, we created our own datasets: Pokemon-Zero for zero overlap and Pokemon-Neg for negative overlap. In this section, we describe the process of dataset creation.