Inference of disease-gene associations helps unravel the pathogenesis of diseases and contributes to the treatment. Although many machine learning-based methods have been developed to predict causative genes, accurate association inference remains challenging. One major reason is the inaccurate feature selection and accumulation of error brought by commonly used multi-stage training architecture. In addition, the existing methods do not incorporate cell-type-specific information, thus fail to study gene functions at a higher resolution. Therefore, we introduce single-cell transcriptome data and construct a context-aware network to unbiasedly integrate all data sources. Then we develop a graph convolution-based approach named CIPHER-SC to realize a complete end-to-end learning architecture. Our approach outperforms four state-of-the-art approaches in five-fold cross-validations on three distinct test sets with the best AUC of 0.9501, demonstrating its stable ability either to predict the novel genes or to predict with genetic basis. The ablation study shows that our complete end-to-end design and unbiased data integration boost the performance from 0.8727 to 0.9443 in AUC. The addition of single-cell data further improves the prediction accuracy and makes our results be enriched for cell-type-specific genes. These results confirm the ability of CIPHER-SC to discover reliable disease genes. Our implementation is available at http://github.com/YidingZhang117/CIPHER-SC.

3D object detection based on LiDAR point cloud has gained more and more attention from industry and academia. Most of the previous works are carried out on the KITTI dataset, which has the gap with real-world scenes in terms of data size and distribution. Faced with large-scale datasets, existing methods have encountered with some problems, especially in the detection of difficult targets and heading prediction. In this paper, we design an integrated framework to improve 3D object detection accuracy targeting on large-scale datasets. We explore data augmentation methods from the perspective of data-rich scenes. Two well-designed auxiliary losses are introduced into our framework to relieve the problem caused by the diversity of points density and distribution. To promote the performance of heading prediction, the existing heading classifier is replaced by our novel dual heading classifier. Experiments on the largest and most realistic 3D object detection benchmark of Waymo Open Dataset show that our integrated framework boosts the detection accuracy for a large margin.

LiDAR-based 3D object detection is important for numerous applications, such as autonomous driving and robot navigation. How to use the specialized geometric representation of point clouds, such as surface normal, to promote the performance of 3D detection is an essential but seldom explored issue. In this paper, we introduce a normal module using geometric surface normal representation to supplement existing methods using coordinate representation. We also propose an attention module for further integration. It integrates features learning from the two representations – coordinate and surface normal, so that they can make up for each other's feature space. Our framework can improve the object detection accuracy, especially for targets with complementary coordinate and surface normal features. As far as we know, we are the first work applying surface normal of LiDAR point cloud to 3D object detection, which can be easily extended to other LiDAR-based methods. Extensive experiments on the KITTI dataset demonstrate that our method boosts the performance of 3D object detection for the voxel-based methods and achieves state-of-the-art results.

Aerial imagery plays an important role in land-use planning, population analysis, precision agriculture, and unmanned aerial vehicle tasks. However, existing aerial image datasets generally suffer from the problem of inaccurate labeling, single ground truth type, and few category numbers. In this work, we implement a simulator that can simultaneously acquire diverse visual ground truth data in the virtual environment. Based on that, we collect a comprehensive Virtual AeriaL Image Dataset named VALID, consisting of 6690 high-resolution images, all annotated with panoptic segmentation on 30 categories, object detection with oriented bounding box, and binocular depth maps, collected in 6 different virtual scenes and 5 various ambient conditions (sunny, dusk, night, snow and fog). To our knowledge, VALID is the first aerial image dataset that can provide panoptic level segmentation and complete dense depth maps. We analyze the characteristics of VALID and evaluate state-of-the-art methods for multiple tasks to provide reference baselines. The experiment results demonstrate that VALID is well presented and challenging. The dataset is available at https://sites.google.com/view/valid-dataset/.

In this paper, we propose a new pipeline of training a monocular UAV to fly a collision-free trajectory along the dense forest trail. As gathering high-precision images in the real world is expensive and the off-the-shelf dataset has some deficiencies, we collect a new dense forest trail dataset in a variety of simulated environment in Unreal Engine. Then we formulate visual perception of forests as a classification problem. A ResNet-18 model is trained to decide the moving direction frame by frame. To transfer the learned strategy to the real world, we construct a ResNet-18 adaptation model via multi-kernel maximum mean discrepancies to leverage the relevant labelled data and alleviate the discrepancy between simulated and real environment. Simulation and real-world flight with a variety of appearance and environment changes are both tested. The ResNet-18 adaptation and its variant model achieve the best result of 84.08% accuracy in reality.