Dynamic graph transformer network via dual-view connectivity for autism spectrum disorder identification.

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder that requires objective and accurate identification methods for effective early intervention. Previous population-based methods via functional connectivity (FC) analysis ignore the differences between positive and negative FCs, which provide the potential information complementarity. And they also require additional information to construct a pre-defined graph. Meanwhile, two challenging demand attentions are the imbalance of performance caused by the class distribution and the inherent heterogeneity of multi-site data. In this paper, we propose a novel dynamic graph Transformer network based on dual-view connectivity for ASD Identification. It is based on the Autoencoders, which regard the input feature as individual feature and without any inductive bias. First, a dual-view feature extractor is designed to extract individual and complementary information from positive and negative connectivity. Then Graph Transformer network is innovated with a hot plugging K-Nearest Neighbor (KNN) algorithm module which constructs a dynamic population graph without any additional information. Additionally, we introduce the PolyLoss function and the Vrex method to address the class imbalance and improve the model's generalizability. The evaluation experiment on 1102 subjects from the ABIDE I dataset demonstrates our method can achieve superior performance over several state-of-the-art methods and satisfying generalizability for ASD identification.

Panicle-Cloud: An Open and AI-Powered Cloud Computing Platform for Quantifying Rice Panicles from Drone-Collected Imagery to Enable the Classification of Yield Production in Rice.

Rice (Oryza sativa) is an essential stable food for many rice consumption nations in the world and, thus, the importance to improve its yield production under global climate changes. To evaluate different rice varieties' yield performance, key yield-related traits such as panicle number per unit area (PNpM2) are key indicators, which have attracted much attention by many plant research groups. Nevertheless, it is still challenging to conduct large-scale screening of rice panicles to quantify the PNpM2 trait due to complex field conditions, a large variation of rice cultivars, and their panicle morphological features. Here, we present Panicle-Cloud, an open and artificial intelligence (AI)-powered cloud computing platform that is capable of quantifying rice panicles from drone-collected imagery. To facilitate the development of AI-powered detection models, we first established an open diverse rice panicle detection dataset that was annotated by a group of rice specialists; then, we integrated several state-of-the-art deep learning models (including a preferred model called Panicle-AI) into the Panicle-Cloud platform, so that nonexpert users could select a pretrained model to detect rice panicles from their own aerial images. We trialed the AI models with images collected at different attitudes and growth stages, through which the right timing and preferred image resolutions for phenotyping rice panicles in the field were identified. Then, we applied the platform in a 2-season rice breeding trial to valid its biological relevance and classified yield production using the platform-derived PNpM2 trait from hundreds of rice varieties. Through correlation analysis between computational analysis and manual scoring, we found that the platform could quantify the PNpM2 trait reliably, based on which yield production was classified with high accuracy. Hence, we trust that our work demonstrates a valuable advance in phenotyping the PNpM2 trait in rice, which provides a useful toolkit to enable rice breeders to screen and select desired rice varieties under field conditions.

MC-UNet: Multimodule Concatenation Based on U-Shape Network for Retinal Blood Vessels Segmentation.

Accurate retinal blood vessels segmentation is an important step in the clinical diagnosis of ophthalmic diseases. Many deep learning frameworks have come up for retinal blood vessels segmentation tasks. However, the complex vascular structure and uncertain pathological features make blood vessel segmentation still very challenging. This paper proposes a novel multimodule concatenation via a U-shaped network for retinal vessels segmentation, which is based on atrous convolution and multikernel pooling. The proposed network structure retains three layers of the essential structure of U-Net, in which the atrous convolution combining the multikernel pooling blocks are designed to obtain more contextual information. The spatial attention module is concatenated with the dense atrous convolution module and the multikernel pooling module to form a multimodule concatenation. And different dilation rates are selected by cascading to acquire a larger receptive field in atrous convolution. Adequate comparative experiments are conducted on these public retinal datasets: DRIVE, STARE, and CHASE_DB1. The results show that the proposed method is effective, especially for microvessels. The code will be released at https://github.com/rocklijun/MC-UNet.

UCR-Net: U-shaped context residual network for medical image segmentation.

Medical image segmentation prerequisite for numerous clinical needs is a critical step in biomedical image analysis. The U-Net framework is one of the most popular deep networks in this field. However, U-Net's successive pooling and downsampling operations result in some loss of spatial information. In this paper, we propose a U-shaped context residual network, called UCR-Net, to capture more context and high-level information for medical image segmentation. The proposed UCR-Net is an encoder-decoder framework comprising a feature encoder module and a feature decoder module. The feature decoder module contains four newly proposed context attention exploration(CAE) modules, a newly proposed global and spatial attention (GSA) module, and four decoder blocks. We use the proposed CAE module to capture more multi-scale context features from the encoder. The proposed GSA module further explores global context features and semantically enhanced deep-level features. The proposed UCR-Net can recover more high-level semantic features and fuse context attention information from CAE and global and spatial attention information from GSA module. Experiments on the retinal vessel, femoropopliteal artery stent, and polyp datasets demonstrate that the proposed UCR-Net performs favorably against the original U-Net and other advanced methods.

MHA-Net: A Multibranch Hybrid Attention Network for Medical Image Segmentation.

The robust segmentation of organs from the medical image is the key technique in medical image analysis for disease diagnosis. U-Net is a robust structure for medical image segmentation. However, U-Net adopts consecutive downsampling encoders to capture multiscale features, resulting in the loss of contextual information and insufficient recovery of high-level semantic features. In this paper, we present a new multibranch hybrid attention network (MHA-Net) to capture more contextual information and high-level semantic features. The main idea of our proposed MHA-Net is to use the multibranch hybrid attention feature decoder to recover more high-level semantic features. The lightweight pyramid split attention (PSA) module is used to connect the encoder and decoder subnetwork to obtain a richer multiscale feature map. We compare the proposed MHA-Net to state-of-art approaches on the DRIVE dataset, the fluoroscopic roentgenographic stereophotogrammetric analysis X-ray dataset, and the polyp dataset. The experimental results on different modal images reveal that our proposed MHA-Net provides better segmentation results than other segmentation approaches.

Learning Two-View Correspondences and Geometry via Local Neighborhood Correlation.

Seeking quality feature correspondences (also known as matches) is a foundational step in computer vision. In our work, a novel and effective network with a stable local constraint, named the Local Neighborhood Correlation Network (LNCNet), is proposed to capture abundant contextual information of each correspondence in the local region, followed by calculating the essential matrix and camera pose estimation. Firstly, the k-Nearest Neighbor (KNN) algorithm is used to divide the local neighborhood roughly. Then, we calculate the local neighborhood correlation matrix (LNC) between the selected correspondence and other correspondences in the local region, which is used to filter outliers to obtain more accurate local neighborhood information. We cluster the filtered information into feature vectors containing richer neighborhood contextual information so that they can be used to more accurately determine the probability of correspondences as inliers. Extensive experiments have demonstrated that our proposed LNCNet performs better than some state-of-the-art networks to accomplish outlier rejection and camera pose estimation tasks in complex outdoor and indoor scenes.

Classification of Rice Yield Using UAV-Based Hyperspectral Imagery and Lodging Feature.

High-yield rice cultivation is an effective way to address the increasing food demand worldwide. Correct classification of high-yield rice is a key step of breeding. However, manual measurements within breeding programs are time consuming and have high cost and low throughput, which limit the application in large-scale field phenotyping. In this study, we developed an accurate large-scale approach and presented the potential usage of hyperspectral data for rice yield measurement using the XGBoost algorithm to speed up the rice breeding process for many breeders. In total, 13 japonica rice lines in regional trials in northern China were divided into different categories according to the manual measurement of yield. Using an Unmanned Aerial Vehicle (UAV) platform equipped with a hyperspectral camera to capture images over multiple time series, a rice yield classification model based on the XGBoost algorithm was proposed. Four comparison experiments were carried out through the intraline test and the interline test considering lodging characteristics at the midmature stage or not. The result revealed that the degree of lodging in the midmature stage was an important feature affecting the classification accuracy of rice. Thus, we developed a low-cost, high-throughput phenotyping and nondestructive method by combining UAV-based hyperspectral measurements and machine learning for estimation of rice yield to improve rice breeding efficiency.

Efficient Robust Model Fitting for Multistructure Data Using Global Greedy Search.

In this paper, a new robust model fitting method is proposed to efficiently segment multistructure data even when they are heavily contaminated by outliers. The proposed method is composed of three steps: first, a conventional greedy search strategy is employed to generate (initial) model hypotheses based on the sequential "fit-and-remove" procedure because of its computational efficiency. Second, to efficiently generate accurate model hypotheses close to the true models, a novel global greedy search strategy initially samples from the inliers of the obtained model hypotheses and samples subsequent data subsets from the whole input data. Third, mutual information theory is applied to fuse the model hypotheses of the same model instance. The conventional greedy search strategy is used to generate model hypotheses for the remaining model instances, if the number of retained model hypotheses is less than that of the true model instances after fusion. The second and the third steps are performed iteratively until an adequate solution is obtained. Experimental results demonstrate the effectiveness and efficiency of the proposed method for model fitting.

Directional support value of Gaussian transformation for infrared small target detection.

Robust small target detection is one of the key techniques in IR search and tracking systems for self-defense or attacks. In this paper we present a robust solution for small target detection in a single IR image. The key ideas of the proposed method are to use the directional support value of Gaussian transform (DSVoGT) to enhance the targets, and use the multiscale representation provided by DSVoGT to reduce the false alarm rate. The original image is decomposed into sub-bands in different orientations by convolving the image with the directional support value filters, which are deduced from the weighted mapped least-squares-support vector machines (LS-SVMs). Based on the sub-band images, a support value of Gaussian matrix is constructed, and the trace of this matrix is then defined as the target measure. The corresponding multiscale correlations of the target measures are computed for enhancing target signal while suppressing the background clutter. We demonstrate the advantages of the proposed method on real IR images and compare the results against those obtained from standard detection approaches, including the top-hat filter, max-mean filter, max-median filter, min-local-Laplacian of Gaussian (LoG) filter, as well as LS-SVM. The experimental results on various cluttered background images show that the proposed method outperforms other detectors.