In this work, we aim to reconstruct the 3D shape of an indoor scene from a single view, which includes multiple objects and the background. This task is challenging for existing methods since those instances of indoor scenes regularly occlude each other and contain diverse topologies. To address this, we propose a novel framework, ISDNet, to adaptively separate mixed instances and perform topology-aware reconstruction. Specifically, Specifically, ISDNet consists of two cascaded subnetworks: an instance separation module (ISM) and an instance deformation module (IDM). The ISM learns to separate occluded objects through stepwise sampling, inferring clean features for each instance. On the basis of these features, IDM generates an instance-topology-aware template and deforms it with learned offsets to reconstruct detailed geometry. Quantitative and qualitative experiments on the SUNRGB-D and 3D-FRONT datasets demonstrate that ISDNet outperforms the state-of-theart methods in terms of local details and overall shapes.

Drug combination synergy is crucial in pharmacology, as it can enhance disease treatment efficacy or reduce drug resistance when administered in combination. Accurate prediction of drug combination synergy is vital for optimizing therapeutic regimens and improving treatment effectiveness. However, existing computational methods primarily rely on drug sequence and structural features, making them difficult to capture complex network relationships and global information-especially lacking the ability to perform cross-modal fusion. In this study, we proposed a method called SDCGDCP for predicting drug combination synergy. It processed drug molecular structures (graph structures and molecular fingerprints), target biological activity information and integrated cell line whole-genome expression profiles to construct multi-level combined node representations. A drug-cell line pair graph was accordingly generated. SDCGDCP updated node and edge representations via a GatedGCN module and derived five types of structural encodings (centrality, spatial, edge, Laplacian positional and node-similarity encodings), after which the graph language model Graphormer is employed to capture long-range node interactions. Finally, drug combination synergy was predicted using an MLP and SoftMax classifier. Extensive evaluations showed that SDCGDCP outperforms other state-of-the-art methods on the DrugCombDB dataset, achieving an AUROC of 0.923 and AUPRC of 0.885. Ablation experiments validated the effectiveness of each feature and encoding module. Meanwhile, we conducted case analyses on the predicted drug combination synergies. The results were supported by evidence from several pharmaceutical studies. This highlights potential of SDCGDCP in enhancing drug synergy prediction and optimizing combination therapies. The source code and data for SDCGDCP are available at https://github.com/Philosopher-Zhao/SDCGDCP.

Automatic delineation of organ-at-risk (OAR) and gross-tumor-volume (GTV) is of great significance for radiotherapy planning. However, it is a challenging task to learn powerful representations for accurate delineation under limited pixel (voxel)-wise annotations. Contrastive learning at pixel-level can alleviate the dependency on annotations by learning dense representations from unlabeled data. Recent studies in this direction design various contrastive losses on the feature maps, to yield discriminative features for each pixel in the map. However, pixels in the same map inevitably share semantics to be closer than they actually are, which may affect the discrimination of pixels in the same map and lead to the unfair comparison to pixels in other maps. To address these issues, we propose a separated region-level contrastive learning scheme, namely SepaReg, the core of which is to separate each image into regions and encode each region separately. Specifically, SepaReg comprises two components: a structure-aware image separation (SIS) module and an intra- and inter-organ distillation (IID) module. The SIS is proposed to operate on the image set to rebuild a region set under the guidance of structural information. The inter-organ representation will be learned from this set via typical contrastive losses cross regions. On the other hand, the IID is proposed to tackle the quantity imbalance in the region set as tiny organs may produce fewer regions, by exploiting intra-organ representations. We conducted extensive experiments to evaluate the proposed model on a public dataset and two private datasets. The experimental results demonstrate the effectiveness of the proposed model, consistently achieving better performance than state-of-the-art approaches. Code is available at https://github.com/jcwang123/Separate_CL.

Non-intrusive sign language recognition has started to redefine the wellness of millions of hearing impaired people around the world. In this study we propose two models, named ContiNet and F-ContiNet. ContiNet is a 3D convolutional network with a constant number of temporal snapshots, and F-ContiNet is an evolution of ContiNet where the fusion idea is applied so that the spatial and temporal information from the early layers will have greater direct impact on the final recognition. We have conducted extensive experiments on a 500-class dataset and its 100-class subset, using both raw videos and mask videos produced by the skeletal information obtained from OpenPose. For all the models we have tested, those using mask videos performed significantly better and learned faster than those using raw videos, and F-ContiNet performed the best, with the best per-class top-1 accuracy of 96.6% over the 500-class dataset. More importantly, it confirms that keeping a constant temporal dimension in a deep network is a viable approach to SLR, especially when the temporal information from all the available snapshots can be fully exploited by a time-sensitive construct such as ConvLSTM.

In heterogeneous distributed cloud systems, load balance, communication and synchronization overhead must be taken considered. A hierarchical distributed loop self-scheduling scheme is effective and efficient for scientific loop applications. In this paper, we propose a resilient hierarchical distributed loop self-scheduling algorithm suitable for heterogeneous cloud systems. This algorithm is intended to enable the algorithm to continue to work in the event that some virtual machines (VMs) are too slow or cease to respond. We tested our algorithm in a heterogeneous cloud system. The results show that our algorithm can achieve normal operation and good performance.

Cloud systems have demonstrated the powerful computation and storage capability in many scientific applications. In this paper, we propose a hierarchical distributed loop self-scheduling scheme to achieve good load balancing by applying weighted self-scheduling scheme on a heterogeneous cloud system. This scheme also considers the distribution of the output data, which can help reduce communication overhead. We evaluated the scheme with two scientific applications: Matrix Multiplication and Quick Sort. The results shows that our schemes achieve better load balancing and better overall performance than standard loop self-scheduling scheme.