Object Identification by Signal Gain and Correlation

In this study, “ring rotation invariant transform” techniques are used to add more salient feature to the original images. The “ring rotation invariant transform” can solve image rotation problem, which transfers a ring signal to several signal vectors in the complex domain, whereby to generate invariant magnitude. Matrix correlation is employed to combine these magnitudes to generate the various discriminators, by which to identify objects. For managing image-shifting problem, one pixel in sample image is compared with surrounding pixels of unknown image. The comparison approaching in this study is by the basis of pixel-to-pixel-comparisons.

Rotation Invariant Parallel Signal Processing Using a Diffractive Phase Element for Image Compression

We propose a new rotation invariant correlator using dimensionality reduction. A diffractive phase element is used to focus image data into a line which serves as input for a conventional correlator. The diffractive element sums information over each radius of the scene image and projects the result onto one point of a line located at a certain distance behind the image. The method is flexible, to a large extent, and might include parallel pattern recognition and classification as well as further geometrical invariance. Although the new technique is inspired from circular harmonic decomposition, it does not suffer from energy loss. A theoretical analysis, as well as examples, are given.

A Novel Method of Aircraft Detection under Complex Background Based on Circular Intensity Filter and Rotation Invariant Feature

Aircraft detection in remote sensing images (RSIs) has drawn widespread attention in recent years, which has been widely used in the military and civilian fields. While the complex background, variations of aircraft pose and size bring great difficulties to the effective detection. In this paper, we propose a novel aircraft target detection scheme based on small training samples. The scheme is coarse-to-fine, which consists of two main stages: region proposal and target identification. First, in the region proposal stage, a circular intensity filter, which is designed based on the characteristics of the aircraft target, can quickly locate the centers of multi-scale suspicious aircraft targets in the RSIs pyramid. Then the target regions can be extracted by adding bounding boxes. This step can get high-quality but few candidate regions. Second, in the stage of target identification, we proposed a novel rotation-invariant feature, which combines rotation-invariant histogram of oriented gradient and vector of locally aggregated descriptors (VLAD). The feature can characterize the aircraft target well by avoiding the impact of its rotation and can be effectively used to remove false alarms. Experiments are conducted on Remote Sensing Object Detection (RSOD) dataset to compare the proposed method with other advanced methods. The results show that the proposed method can quickly and accurately detect aircraft targets in RSIs and achieve a better performance.

DeepRank-GNN: A Graph Neural Network Framework to Learn Patterns in Protein-Protein Interfaces

Gaining structural insights into the protein-protein interactome is essential to understand biological phenomena and extract knowledge for rational drug design or protein engineering. We have previously developed DeepRank, a deep-learning framework to facilitate pattern learning from protein-protein interfaces using Convolutional Neural Network (CNN) approaches. However, CNN is not rotation invariant and data augmentation is required to desensitize the network to the input data orientation which dramatically impairs the computation performance. Representing protein-protein complexes as atomic- or residue-scale rotation invariant graphs instead enables using graph neural networks (GNN) approaches, bypassing those limitations. We have developed DeepRank-GNN, a framework that converts protein-protein interfaces from PDB 3D coordinates files into graphs that are further provided to a pre-defined or user-defined GNN architecture to learn problem-specific interaction patterns. DeepRank-GNN is designed to be highly modularizable, easily customized, and is wrapped into a user-friendly python3 package. Here, we showcase DeepRank-GNN's performance for scoring docking models using a dedicated graph interaction neural network (GINet). We show that this graph-based model performs better than DeepRank, DOVE and HADDOCK scores and competes with iScore on the CAPRI score set. We show a significant gain in speed and storage requirement using DeepRank-GNN as compared to DeepRank. DeepRank-GNN is freely available from https://github.com/DeepRank/DeepRank-GNN.

An Efficient and Robust Star Identification Algorithm Based on Neural Networks

A lost-in-space star identification algorithm based on a one-dimensional Convolutional Neural Network (1D CNN) is proposed. The lost-in-space star identification aims to identify stars observed with corresponding catalog stars when there is no prior attitude information. With the help of neural networks, the robustness and the speed of the star identification are improved greatly. In this paper, a modified log-Polar mapping is used to constructed rotation-invariant star patterns. Then a 1D CNN is utilized to classify the star patterns associated with guide stars. In the 1D CNN model, a global average pooling layer is used to replace fully-connected layers to reduce the number of parameters and the risk of overfitting. Experiments show that the proposed algorithm is highly robust to position noise, magnitude noise, and false stars. The identification accuracy is 98.1% with 5 pixels position noise, 97.4% with 5 false stars, and 97.7% with 0.5 Mv magnitude noise, respectively, which is significantly higher than the identification rate of the pyramid, optimized grid and modified log-polar algorithms. Moreover, the proposed algorithm guarantees a reliable star identification under dynamic conditions. The identification accuracy is 82.1% with angular velocity of 10 degrees per second. Furthermore, its identification time is as short as 32.7 miliseconds and the memory required is about 1920 kilobytes. The algorithm proposed is suitable for current embedded systems.

Rotation-Invariant and Relation-Aware Cross-Domain Adaptation Object Detection Network for Optical Remote Sensing Images

In recent years, object detection has shown excellent results on a large number of annotated data, but when there is a discrepancy between the annotated data and the real test data, the performance of the trained object detection model is often degraded when it is directly transferred to the real test dataset. Compared with natural images, remote sensing images have great differences in appearance and quality. Traditional methods need to re-label all image data before interpretation, which will consume a lot of manpower and time. Therefore, it is of practical significance to study the Cross-Domain Adaptation Object Detection (CDAOD) of remote sensing images. To solve the above problems, our paper proposes a Rotation-Invariant and Relation-Aware (RIRA) CDAOD network. We trained the network at the image-level and the prototype-level based on a relation aware graph to align the feature distribution and added the rotation-invariant regularizer to deal with the rotation diversity. The Faster R-CNN network was adopted as the backbone framework of the network. We conducted experiments on two typical remote sensing building detection datasets, and set three domain adaptation scenarios: WHU 2012 → WHU 2016, Inria (Chicago) → Inria (Austin), and WHU 2012 → Inria (Austin). [d=Y.C.]The results show that our method can effectively improve the detection effect in the target domain, and outperform competing methods by obtaining optimal results in all three scenarios.The results show that our method can effectively improve the detection effect in the target domain, and in the comparison methods, we get the optimal results in all three scenarios.