GLPanoDepth: Global-to-Local Panoramic Depth Estimation.

Depth estimation is a fundamental task in many vision applications. With the popularity of omnidirectional cameras, it becomes a new trend to tackle this problem in the spherical space. In this paper, we propose a learning-based method for predicting dense depth values of a scene from a monocular omnidirectional image. An omnidirectional image has a full field-of-view, providing much more complete descriptions of the scene than perspective images. However, fully-convolutional networks that most current solutions rely on fail to capture rich global contexts from the panorama. To address this issue and also the distortion of equirectangular projection in the panorama, we propose Cubemap Vision Transformers (CViT), a new transformer-based architecture that can model long-range dependencies and extract distortion-free global features from the panorama. We show that cubemap vision transformers have a global receptive field at every stage and can provide globally coherent predictions for spherical signals. As a general architecture, it removes any restriction that has been imposed on the panorama in many other monocular panoramic depth estimation methods. To preserve important local features, we further design a convolution-based branch in our pipeline (dubbed GLPanoDepth) and fuse global features from cubemap vision transformers at multiple scales. This global-to-local strategy allows us to fully exploit useful global and local features in the panorama, achieving state-of-the-art performance in panoramic depth estimation.

Low-Velocity Impact and Post-Impact Residual Flexural Properties of Kevlar/EP Three-Dimensional Angle-Interlock Composites.

In this study, five three-dimensional angle-interlock fabrics with different warp and weft densities were fabricated using 1000D Kevlar filaments. The Kevlar/EP composites were prepared by vacuum-assisted molding techniques. The low-velocity impact property of the composite was tested, focusing on the effects of the warp and weft densities, impact energy, impactor shape, and impactor diameter. The damage area, dent depth, and crack lengths in the warp and weft direction were used to evaluate the impact performance, and the specimens were compared with plain-weave composites with similar areal densities. The dominant failure mode of the conical impactor was fiber fracture, while the dominant failure mode of the hemispherical impactor was fiber-resin debonding. The cylindrical impactor showed only minor resin fragmentation. The residual flexural strength of the composite after impact was tested to provide insights into its mechanical properties. The study findings will provide a theoretical basis for the optimization of the design of impact-resistant structures using such materials and facilitate their engineering applications.

Environmental impact assessment of acidic coal gangue leaching solution on groundwater: a coal gangue pile in Shanxi, China.

With the continual advancement of coal resource development, the comprehensive utilization of coal gangue as a by-product encounters certain constraints. A substantial amount of untreated coal gangue is openly stored, particularly acidic gangue exposed to rainfall. The leaching effect of acidic solutions, containing heavy metal ions and other pollutants, results in environmental challenges such as local soil or groundwater pollution, presenting a significant concern in the current ecological landscape of mining areas. Investigating the migration patterns of pollutants in the soil-groundwater system and elucidating the characteristics of polluted solute migration are imperative. To understand the migration dynamics of pollutants and unveil the features of solute migration, this study focuses on a coal gangue dump in a mining area in Shanxi. Utilizing indoor leaching experiments and soil column migration experiments, a two-dimensional soil-groundwater model is established using the finite element method of COMSOL. This model quantitatively delineates the migration patterns of key pollutant components leached from coal gangue into the groundwater. The findings reveal that sulfate ions can migrate and infiltrate groundwater within a mere 7 years in the vadose zone of aeration. Moreover, the average concentration of iron ions in groundwater can reach approximately 58.3 mg/L. Convection, hydrodynamic dispersion, and adsorption emerge as the primary factors influencing pollution transport. Understanding the leaching patterns and environmental impacts of major pollutants in acidic coal gangue is crucial for predicting soil-groundwater pollution and implementing effective protective measures.

Collaborative Completion and Segmentation for Partial Point Clouds with Outliers.

Outliers will inevitably creep into the captured point cloud during 3D scanning, degrading cutting-edge models on various geometric tasks heavily. This paper looks at an intriguing question that whether point cloud completion and segmentation can promote each other to defeat outliers. To answer it, we propose a collaborative completion and segmentation network, termed CS-Net, for partial point clouds with outliers. Unlike most of existing methods, CS-Net does not need any clean (or say outlier-free) point cloud as input or any outlier removal operation. CS-Net is a new learning paradigm that makes completion and segmentation networks work collaboratively. With a cascaded architecture, our method refines the prediction progressively. Specifically, after the segmentation network, a cleaner point cloud is fed into the completion network. We design a novel completion network which harnesses the labels obtained by segmentation together with farthest point sampling to purify the point cloud and leverages KNN-grouping for better generation. Benefited from segmentation, the completion module can utilize the filtered point cloud which is cleaner for completion. Meanwhile, the segmentation module is able to distinguish outliers from target objects more accurately with the help of the clean and complete shape inferred by completion. Besides the designed collaborative mechanism of CS-Net, we establish a benchmark dataset of partial point clouds with outliers. Extensive experiments show clear improvements of our CS-Net over its competitors, in terms of outlier robustness and completion accuracy.

Local-to-Global Panorama Inpainting for Locale-Aware Indoor Lighting Prediction.

Predicting panoramic indoor lighting from a single perspective image is a fundamental but highly ill-posed problem in computer vision and graphics. To achieve locale-aware and robust prediction, this problem can be decomposed into three sub-tasks: depth-based image warping, panorama inpainting and high-dynamic-range (HDR) reconstruction, among which the success of panorama inpainting plays a key role. Recent methods mostly rely on convolutional neural networks (CNNs) to fill the missing contents in the warped panorama. However, they usually achieve suboptimal performance since the missing contents occupy a very large portion in the panoramic space while CNNs are plagued by limited receptive fields. The spatially-varying distortion in the spherical signals further increases the difficulty for conventional CNNs. To address these issues, we propose a local-to-global strategy for large-scale panorama inpainting. In our method, a depth-guided local inpainting is first applied on the warped panorama to fill small but dense holes. Then, a transformer-based network, dubbed PanoTransformer, is designed to hallucinate reasonable global structures in the large holes. To avoid distortion, we further employ cubemap projection in our design of PanoTransformer. The high-quality panorama recovered at any locale helps us to capture spatially-varying indoor illumination with physically-plausible global structures and fine details.

Precise Electron-Withdrawing Strength Modulation of ESIPT Probes for Ultrasensitive and Specific Fluorescence Sensing.

The precise regulation of the electron-withdrawing/electron-donating strength in a probe is of great significance for the design of reaction-based fluorescent probes with specific functionalities. Here, a family of excited-state intramolecular proton transfer (ESIPT)-based probes with fluorescence turn-on sensing properties toward KMnO4 was designed by precisely modulating the electron-withdrawing strength of the substituents located at the para-position of the recognition group. It is found that -F, -CHO, and -H as the electron-withdrawing groups bound at the probe can specifically recognize KMnO4, which ensures a blue emission displayed by the reaction products. Especially with -CHO as the electron-withdrawing group, the reaction product shows the most stable fluorescence. The probe 2-(benzo[d]oxazol-2-yl)-4-formylphenyl acrylate (BOPA-CHO) demonstrated a more superior sensing performance toward KMnO4, including a low limit of detection (LOD, 0.96 nM), a rapid response (<3 s), and a rather good selectivity even in the presence of 21 interferents. Moreover, the practicality of the probe was further verified by a test pen comprising a BOPA-CHO-embedded sponge, which is capable of detecting KMnO4 solid with a naked-eye LOD of 11.62 ng. The present probe design and modulation strategy would open up a new path for the design of high-performance fluorescent probes.

ShadowMover: Automatically Projecting Real Shadows onto Virtual Object.

Inserting 3D virtual objects into real-world images has many applications in photo editing and augmented reality. One key issue to ensure the reality of the composite whole scene is to generate consistent shadows between virtual and real objects. However, it is challenging to synthesize visually realistic shadows for virtual and real objects without any explicit geometric information of the real scene or manual intervention, especially for the shadows on the virtual objects projected by real objects. In view of this challenge, we present, to our knowledge, the first end-to-end solution to fully automatically project real shadows onto virtual objects for outdoor scenes. In our method, we introduce the Shifted Shadow Map, a new shadow representation that encodes the binary mask of shifted real shadows after inserting virtual objects in an image. Based on the shifted shadow map, we propose a CNN-based shadow generation model named ShadowMover which first predicts the shifted shadow map for an input image and then automatically generates plausible shadows on any inserted virtual object. A large-scale dataset is constructed to train the model. Our ShadowMover is robust to various scene configurations without relying on any geometric information of the real scene and is free of manual intervention. Extensive experiments validate the effectiveness of our method.

AGConv: Adaptive Graph Convolution on 3D Point Clouds.

Convolution on 3D point clouds is widely researched yet far from perfect in geometric deep learning. The traditional wisdom of convolution characterises feature correspondences indistinguishably among 3D points, arising an intrinsic limitation of poor distinctive feature learning. In this article, we propose Adaptive Graph Convolution (AGConv) for wide applications of point cloud analysis. AGConv generates adaptive kernels for points according to their dynamically learned features. Compared with the solution of using fixed/isotropic kernels, AGConv improves the flexibility of point cloud convolutions, effectively and precisely capturing the diverse relations between points from different semantic parts. Unlike the popular attentional weight schemes, AGConv implements the adaptiveness inside the convolution operation instead of simply assigning different weights to the neighboring points. Extensive evaluations clearly show that our method outperforms state-of-the-arts of point cloud classification and segmentation on various benchmark datasets. Meanwhile, AGConv can flexibly serve more point cloud analysis approaches to boost their performance. To validate its flexibility and effectiveness, we explore AGConv-based paradigms of completion, denoising, upsampling, registration and circle extraction, which are comparable or even superior to their competitors.

GeoDualCNN: Geometry-Supporting Dual Convolutional Neural Network for Noisy Point Clouds.

We propose a geometry-supporting dual convolutional neural network (GeoDualCNN) for both point cloud normal estimation and denoising. GeoDualCNN fuses the geometry domain knowledge that the underlying surface of a noisy point cloud is piecewisely smooth with the fact that a point normal is properly defined only when local surface smoothness is guaranteed. Centered around this insight, we define the homogeneous neighborhood (HoNe) which stays clear of surface discontinuities, and associate each HoNe with a point whose geometry and normal orientation is mostly consistent with that of HoNe. Thus, we not only obtain initial estimates of the point normals by performing PCA on HoNes, but also for the first time optimize these initial point normals by learning the mapping from two proposed geometric descriptors to the ground-truth point normals. GeoDualCNN consists of two parallel branches that remove noise using the first geometric descriptor (a homogeneous height map, which encodes the point-position information), while preserving surface features using the second geometric descriptor (a homogeneous normal map, which encodes the point-normal information). Such geometry-supporting network architectures enable our model to leverage previous geometry expertise and to benefit from training data. Experiments with noisy point clouds show that GeoDualCNN outperforms the state-of-the-art methods in terms of both noise-robustness and feature preservation.

Learning Graph Convolutional Networks for Multi-Label Recognition and Applications.

The task of multi-label image recognition is to predict a set of object labels that present in an image. As objects normally co-occur in an image, it is desirable to model the label dependencies to improve the recognition performance. To capture and explore such important information, we propose graph convolutional networks (GCNs) based models for multi-label image recognition, where directed graphs are constructed over classes and information is propagated between classes to learn inter-dependent class-level representations. Following this idea, we design two particular models that approach multi-label classification from different views. In our first model, the prior knowledge about the class dependencies is integrated into classifier learning. Specifically, we propose Classifier Learning GCN (C-GCN) to map class-level semantic representations (e.g., word embeddings) into classifiers that maintain the inter-class topology. In our second model, we decompose the visual representation of an image into a set of label-aware features and propose prediction learning GCN (P-GCN) to encode such features into inter-dependent image-level prediction scores. Furthermore, we also present an effective correlation matrix construction approach to capture inter-class relationships and consequently guide information propagation among classes. Empirical results on generic multi-label image recognition demonstrate that both of the proposed models can obviously outperform other existing state-of-the-arts. Moreover, the proposed methods also show advantages in some other multi-label classification related applications.

Treatment of fluorinated wastewater with chitosan modified activated sludge lysis ash.

Preparation of a novel environmentally friendly and cost-effective composite adsorbent for fluoride removal is presented in this work. An activated sludge lysis ash/chitosan (ASLA/C) composite adsorbent was synthesised using an in situ coprecipitation method, and the removal effect of the material was analysed by static adsorption, isothermal adsorption and kinetic adsorption tests. Langmuir model could better describe the adsorption process and the adsorption was in accordance with the kinetic equation of the pseudo-second-order kinetics reaction. The values of adsorption thermodynamic and kinetic parameters were indicated that the adsorption of fluoride ions is a spontaneous, heat-absorbing entropic process, and the reaction was carried out by a combination of mechanisms, such as electrostatic adsorption, ion exchange, surface complexation and hydrogen bonding. The experimental results indicated that ASLA/C can be used as a cheap and readily available alternative efficient adsorbent where the maximum fluorinate absorption was observed with 7.714 mg g-1, while solving the problem of waste from activated sludge lysis disposal and realizing the environmental benefits of waste.

Low BMI is associated with poor IUI outcomes: a retrospective study in 13,745 cycles.

PURPOSE: To evaluate the association between body mass index (BMI) and pregnancy outcomes in women receiving intrauterine insemination (IUI) treatment. METHODS: The study included 6407 women undergoing 13,745 IUI cycles stratified by BMI. Cox regression was used to analyze the association between BMI and cumulative live births across multiple IUI cycles. A generalized estimating equation (GEE) was used to analyze the live birth rate per cycle. RESULTS: Compared with normal-weight women (n = 4563), underweight women (n = 990) had a lower cumulative pregnancy and live birth rate (20.71% vs 25.93% and17.17% vs 21.61%, respectively), while overweight women (n = 854) had a higher cumulative pregnancy and live birth rate (31.97%, 26.58%). Adjusted for confounders, the hazard ratio (HR) for achieving live birth following up to a maximum of four IUI cycles was 0.80 (95% CI: 0.67-0.95), comparing underweight with normal weight. In the GEE analyses, low BMI was also associated with a lower per-cycle birth rate (OR 0.79, 95% CI: 0.66-0.95), with adjustment for cycle-specific parameters, including ovarian stimulation, endometrial thickness, and follicular diameter. CONCLUSION: Low BMI is associated with poor IUI outcomes.

Application target image of a small modular fluoride salt-cooled high temperature reactor based on the analysis of energy demand and environmental conditions in western China.

To meet the energy demand in remote areas of western China, a small modular fluoride-salt-cooled high-temperature reactor will be deployed there. The design of this reactor needs an application target image, including the power, lifetime, size, weight, and environmental restrictions. This paper analyzes the energy demand in northwest China to find the possible application scenarios and regions of this reactor. Then according to typical application scenarios, the power requirements of the reactor and the environment and transportation (size and weight) restrictions that need to be adapted can be obtained. The application target image of this reactor would be: a) the single unit capacity of 50MWe; b) the lifetime of at least 60 years; c) the length of less than 15.5 m, the diameter of less than 3.88 m, and weight of less than 60 tons; d) the adaptability to the severe climate and environment in the west, such as cold winter, deep-frozen soil layer, strong wind, arid environment, and complex terrain; e) the site selection to avoid geological disasters, such as landslides, collapses, torrents, mudslides, and ground fissures.

Experimental study on the treatment of acid mine drainage containing heavy metals with domestic waste pyrolysis ash.

Acid mine drainage (AMD) is a special kind of acidic wastewater produced in the process of mining and utilization. In this study, AMD was treated using the adsorption method. Domestic waste was prepared by pyrolysis, and the resulting waste pyrolysis ash adsorbent was studied experimentally by a static adsorption test to treat metal ions in AMD. The results showed that the maximum adsorption amounts of Zn2+, Cu2+, Mn2+, Fe2+, Pb2+, and Cd2+ reached 0.425, 0.593, 0.498, 18.519, 0.055, and 0.039 mg/g, respectively, when the amount of pyrolysis ash was added at 30 g/L, the initial pH of the water was 4.1 and the reaction time was 150 min. It was found that the waste pyrolysis ash could be reused at least three times by using Na2S as the regeneration agent. The SEM and BET characterization results prove that its large specific surface areas and well-developed pore structures have the potential to promote the adsorption of metal ions. The pseudo-second-order kinetic equation and Freundlich adsorption isotherms fit the adsorption process well, and the experiments reveal that the metal ions in AMD are well treated by waste pyrolysis ash through adsorption, flocculation and chemical precipitation. Waste pyrolysis ash has great potential for the treatment of acid mine drainage, providing a new approach to solid waste disposal and new ideas for water treatment as a low-cost alternative material.

Video Vectorization via Bipartite Diffusion Curves Propagation and Optimization.

We propose a new video vectorization approach for converting videos in the raster format to vector representation with the benefits of resolution independence and compact storage. Through classifying extracted curves in each video frame into salient ones and non-salient ones, we introduce a novel bipartite diffusion curves (BDCs) representation in order to preserve both important image features such as sharp boundaries and regions with smooth color variation. This bipartite representation allows us to propagate non-salient curves across frames such that the propagation, in conjunction with geometry optimization and color optimization of salient curves, ensures the preservation of fine details within each frame and across different frames, and meanwhile, achieves good spatial-temporal coherence. Thorough experiments on a variety of videos show that our method is capable of converting videos to the vector representation with low reconstruction errors, low computational cost, and fine details, demonstrating our superior performance over the state of the art. We also show that, when used for video upsampling, our method produces results comparable to video super-resolution.

Semi-Supervised Pixel-Level Scene Text Segmentation by Mutually Guided Network.

In this paper we present a new data-driven method for pixel-level scene text segmentation from a single natural image. Although scene text detection, i.e. producing a text region mask, has been well studied in the past decade, pixel-level text segmentation is still an open problem due to the lack of massive pixel-level labeled data for supervised training. To tackle this issue, we incorporate text region mask as an auxiliary data into this task, considering acquiring large-scale of labeled text region mask is commonly less expensive and time-consuming. To be specific, we propose a mutually guided network which produces a polygon-level mask in one branch and a pixel-level text mask in the other. The two branches' outputs serve as guidance for each other and the whole network is trained via a semi-supervised learning strategy. Extensive experiments are conducted to demonstrate the effectiveness of our mutually guided network, and experimental results show our network outperforms the state-of-the-art in pixel-level scene text segmentation. We also demonstrate the mask produced by our network could improve the text recognition performance besides the trivial image editing application.

HCE: Hierarchical Context Embedding for Region-Based Object Detection.

State-of-the-art two-stage object detectors apply a classifier to a sparse set of object proposals, relying on region-wise features extracted by RoIPool or RoIAlign as inputs. The region-wise features, in spite of aligning well with the proposal locations, may still lack the crucial context information which is necessary for filtering out noisy background detections, as well as recognizing objects possessing no distinctive appearances. To address this issue, we present a simple but effective Hierarchical Context Embedding (HCE) framework, which can be applied as a plug-and-play component, to facilitate the classification ability of a series of region-based detectors by mining contextual cues. Specifically, to advance the recognition of context-dependent object categories, we propose an image-level categorical embedding module which leverages the holistic image-level context to learn object-level concepts. Then, novel RoI features are generated by exploiting hierarchically embedded context information beneath both whole images and interested regions, which are also complementary to conventional RoI features. Moreover, to make full use of our hierarchical contextual RoI features, we propose the early-and-late fusion strategies (i.e., feature fusion and confidence fusion), which can be combined to boost the classification accuracy of region-based detectors. Comprehensive experiments demonstrate that our HCE framework is flexible and generalizable, leading to significant and consistent improvements upon various region-based detectors, including FPN, Cascade R-CNN, Mask R-CNN and PA-FPN. With simple modification, our HCE framework can be conveniently adapted to fit the structure of one-stage detectors, and achieve improved performance for SSD, RetinaNet and EfficientDet.

Image Stitching Based on Semantic Planar Region Consensus.

Image stitching for two images without a global transformation between them is notoriously difficult. In this paper, noticing the importance of semantic planar structures under perspective geometry, we propose a new image stitching method which stitches images by allowing for the alignment of a set of matched dominant semantic planar regions. Clearly different from previous methods resorting to plane segmentation, the key to our approach is to utilize rich semantic information directly from RGB images to extract semantic planar image regions with a deep Convolutional Neural Network (CNN). We specifically design a module implementing our newly proposed clustering loss to make full use of existing semantic segmentation networks to accommodate region segmentation. To train the network, a dataset for semantic planar region segmentation is constructed. With the prior of semantic planar region, a set of local transformation models can be obtained by constraining matched regions, enabling more precise alignment in the overlapping area. We also use this prior to estimate a transformation field over the whole image. The final mosaic is obtained by mesh-based optimization which maintains high alignment accuracy and relaxes similarity transformation at the same time. Extensive experiments with both qualitative and quantitative comparisons show that our method can deal with different situations and outperforms the state-of-the-arts on challenging scenes.

Transferable Multilevel Attention Neural Network for Accurate Prediction of Quantum Chemistry Properties via Multitask Learning.

The development of efficient models for predicting specific properties through machine learning is of great importance for the innovation of chemistry and material science. However, predicting global electronic structure properties like Frontier molecular orbital highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels and their HOMO-LUMO gaps from the small-sized molecule data to larger molecules remains a challenge. Here, we develop a multilevel attention neural network, named DeepMoleNet, to enable chemical interpretable insights being fused into multitask learning through (1) weighting contributions from various atoms and (2) taking the atom-centered symmetry functions (ACSFs) as the teacher descriptor. The efficient prediction of 12 properties including dipole moment, HOMO, and Gibbs free energy within chemical accuracy is achieved by using multiple benchmarks, both at the equilibrium and nonequilibrium geometries, including up to 110,000 records of data in QM9, 400,000 records in MD17, and 280,000 records in ANI-1ccx for random split evaluation. The good transferability for predicting larger molecules outside the training set is demonstrated in both equilibrium QM9 and Alchemy data sets at the density functional theory (DFT) level. Additional tests on nonequilibrium molecular conformations from DFT-based MD17 data set and ANI-1ccx data set with coupled cluster accuracy as well as the public test sets of singlet fission molecules, biomolecules, long oligomers, and protein with up to 140 atoms show reasonable predictions for thermodynamics and electronic structure properties. The proposed multilevel attention neural network is applicable to high-throughput screening of numerous chemical species in both equilibrium and nonequilibrium molecular spaces to accelerate rational designs of drug-like molecules, material candidates, and chemical reactions.

Optimization study on deep extractive oxidative desulfurization with tetrabutylammonium bromide/polyethylene glycol DES.

Green, efficient and inexpensive desulfurizing solvents have always been a considerable focus of petroleum desulfurization research. In this study, a series of deep eutectic solvents (DESs) based on tetrabutylammonium bromide (TBAB)/polyethylene glycol 200 (PEG-200) with different molar ratios were synthesized and characterized by Fourier transform infrared spectroscopy and 1H nuclear magnetic resonance spectroscopy. Dibenzothiophene (DBT) was removed deeply as the classic sulfide in model oil, and H2O2 was fully utilized by the new TBAB/PEG-200 desulfurization system in step extractive oxidative desulfurization. The reaction conditions were optimized further, and O/S = 8, DES/oil = 1 : 5, 40 °C and 75 minutes were chosen as the best reaction conditions. Meanwhile, other organic sulfides in crude oil were also removed, and the removal rates of DBT, 4,6-dimethyldibenzothiophene and benzothiophene were 99.65%, 96.71% and 93.52%, respectively. The DES was reused 7 times, and the desulfurization efficiency of the regenerated DES for DBT was maintained at 98.14%. Finally, the possible mechanism of the synergistic effect of two kinds of hydrogen bonds and the oxidant was proposed.