Search

As an essential protection mechanism of information security, hash algorithms are extensively used in various security mechanisms. The diverse application scenarios make the implementation of hash algorithms more challenging regarding flexibility, performance, and resources. Since the existing studies have such issues as wasted resources and few algorithms are supported when implementing hash algorithms, we proposed a new reconfigurable hardware architecture for common hash algorithms in this paper. First, we used the characteristics of symmetry of SM3 (Shang Mi 3) and SHA2 (Secure Hash Algorithm 2) to design an architecture that also supports MD5 (Message Digest 5) and SHA1 (Secure Hash Algorithm 1) on both sides. Then we split this architecture into two layers and eliminated the resource wastes introduced by different word widths through exploiting greater parallelism. Last, we further divided the architecture into four operators and designed an array. The experimental results showed that our architecture can support four types of hash algorithms successfully, and supports 32-bit and 64-bit word widths without wasting resources. Compared with existing designs, our design has a throughput rate improvement of about 56.87–226% and a throughput rate per resource improvement of up to 5.5 times. Furthermore, the resource utilization rose to 80% or above when executing algorithms.

We study the multilayer random dot product graph (MRDPG) model, an extension of the random dot product graph to multilayer networks. By modelling a multilayer network as an MRDPG, we deploy a tensor-based method and demonstrate its superiority over existing approaches. Moving to dynamic MRDPGs, we focus on online change point detection problems. At every time point, we observe a realisation from an MRDPG. Across layers, we assume shared common node sets and latent positions but allow for different connectivity matrices. We propose efficient algorithms for both fixed and random latent position cases, minimising detection delay while controlling false alarms. Notably, in the random latent position case, we devise a novel nonparametric change point detection algorithm with a kernel estimator in its core, allowing for the case when the density does not exist, accommodating stochastic block models as special cases. Our theoretical findings are supported by extensive numerical experiments, with the code available online https://github.com/MountLee/MRDPG.

Effective control of nonlinear processes at high power levels in multimode fibers (MMFs) would unlock unprecedented possibilities for diverse applications including high-power fiber lasers, bioimaging, chemical sensing, and novel physics phenomena. Existing studies have focused on spatial control of nonlinear effects in graded-index MMFs, limiting their capabilities due to two major challenges: difficulty in control and limited broadband spectral brilliance. Here we present a new control approach that exploits the spatial and temporal degrees of control in step-index MMFs using a 3D-printed programmable fiber shaper. We have achieved broadband high-peak-power spanning 560--2200 nm, resulting from combined spectral energy reallocation (up to 166-fold) and temporal shortening (up to 4-fold) uniquely enabled by the fiber shaper. Our simple but effective fiber shaper costs 35 dollars, making it a potentially accessible tool for nonlinear and linear modulation of MMFs, with important implications in nonlinear optics, bioimaging, spectroscopy, optical computing, and material processing., 31 pages, 12 figures

We investigated a viscous protein layer formed on self-assembled monolayers (SAMs) in crowded biological environments. The results were obtained through force spectroscopic measurements using colloidal probes and substantiated by exhaustive analysis using a quartz crystal microbalance with an energy dissipation technique. A hydrophobic SAM of

Signal region detection is one of the challenging problems in modern statistics and has broad applications especially in genetic studies. We propose a novel approach effectively coupling with high-dimensional test, which is distinct from existing methods based on scan or knockoff statistics. The idea is to conduct binary segmentation with re-search and arrangement based on a sequence of dynamic tests to increase detection accuracy and reduce computation. Theoretical and empirical studies demonstrate that our approach enjoys favorable theoretical guarantees with fewer restrictions and exhibits superior numerical performance with faster computation. Compared to scan-based methods, our procedure is capable of detecting shorter or longer regions with unbalanced signal strengths while allowing for more dependence structures. Relative to the knockoff framework that only controls false discovery rate, our approach attains higher detection accuracy while controlling the family-wise error rate. Utilizing the UK Biobank data to identify the genetic regions related to breast cancer, we confirm previous findings and meanwhile, identify a number of new regions which suggest strong association with risk of breast cancer and deserve further investigation.

An electric orchard tractor with a power battery and transmission driven by dual motors was developed. The output shafts of the walking and power take-off (PTO) motors are connected by a wet clutch, which controls whether the two motors are coupled or independent. When the load of the walking or PTO motor exceeds its output torque, the two motors are driven by power coupled by the wet clutch to meet the power demand. According to the heavy-load working conditions of ploughing and rototilling, the power battery capacity and working duration targets of 15 kW and 4 hours/charge were set. A prototype 15 kW electric orchard tractor was manufactured and assembled, and its performance was assessed. A bench test showed that the tractor’s maximum PTO output power was 13.9 kW and a field rototilling test showed that its maximum continuous working time was 4.5 hours. Thus, the prototype electric orchard tractor met the design goals and requirements of for orchard operations.

Vision Transformers have shown great potential in computer vision tasks. Most recent works have focused on elaborating the spatial token mixer for performance gains. However, we observe that a well-designed general architecture can significantly improve the performance of the entire backbone, regardless of which spatial token mixer is equipped. In this paper, we propose UniNeXt, an improved general architecture for the vision backbone. To verify its effectiveness, we instantiate the spatial token mixer with various typical and modern designs, including both convolution and attention modules. Compared with the architecture in which they are first proposed, our UniNeXt architecture can steadily boost the performance of all the spatial token mixers, and narrows the performance gap among them. Surprisingly, our UniNeXt equipped with naive local window attention even outperforms the previous state-of-the-art. Interestingly, the ranking of these spatial token mixers also changes under our UniNeXt, suggesting that an excellent spatial token mixer may be stifled due to a suboptimal general architecture, which further shows the importance of the study on the general architecture of vision backbone. All models and codes will be publicly available.

BackgroundMethadone maintenance treatment (MMT) is a common treatment for heroin use disorder (HUD). Although individuals with HUD have been reported to show impaired coupling among the salience network (SN), executive control network (ECN), and default mode network (DMN), the effects of MMT on the coupling among three large-scale networks in individuals with HUD remains unclear.MethodsThirty-seven individuals with HUD undergoing MMT and 57 healthy controls were recruited. The longitudinal one-year follow-up study aimed to evaluate the effects of methadone on anxiety, depression, withdrawal symptoms and craving and number of relapse, and brain function (SN, DMN and bilateral ECN) in relation to heroin dependence. The changes in psychological characteristics and the coupling among large-scale networks after 1 year of MMT were analyzed. The associations between the changes in coupling among large-scale networks and psychological characteristics and the methadone dose were also examined.ResultsAfter 1 year of MMT, individuals with HUD showed a reduction in the withdrawal symptom score. The number of relapses was negatively correlated with the methadone dose over 1 year. The functional connectivity between the medial prefrontal cortex (mPFC) and the left middle temporal gyrus (MTG; both key nodes of the DMN) was increased, and the connectivities between the mPFC and the anterior insular and middle frontal gyrus (key nodes of the SN) were also increased. The mPFC-left MTG connectivity was negatively correlated with the withdrawal symptom score.ConclusionLong-term MMT enhanced the connectivity within the DMN which might be related to reduced withdrawal symptoms, and that between the DMN and SN which might be related to increase in salience values of heroin cues in individuals with HUD. Long-term MMT may be a double-edged sword in treatment for HUD.

This paper introduces Amazon Robotic Manipulation Benchmark (ARMBench), a large-scale, object-centric benchmark dataset for robotic manipulation in the context of a warehouse. Automation of operations in modern warehouses requires a robotic manipulator to deal with a wide variety of objects, unstructured storage, and dynamically changing inventory. Such settings pose challenges in perceiving the identity, physical characteristics, and state of objects during manipulation. Existing datasets for robotic manipulation consider a limited set of objects or utilize 3D models to generate synthetic scenes with limitation in capturing the variety of object properties, clutter, and interactions. We present a large-scale dataset collected in an Amazon warehouse using a robotic manipulator performing object singulation from containers with heterogeneous contents. ARMBench contains images, videos, and metadata that corresponds to 235K+ pick-and-place activities on 190K+ unique objects. The data is captured at different stages of manipulation, i.e., pre-pick, during transfer, and after placement. Benchmark tasks are proposed by virtue of high-quality annotations and baseline performance evaluation are presented on three visual perception challenges, namely 1) object segmentation in clutter, 2) object identification, and 3) defect detection. ARMBench can be accessed at http://armbench.com, To appear at the IEEE Conference on Robotics and Automation (ICRA), 2023

Recent advances in digitization have led to the availability of multivariate time series data in various domains, enabling real-time monitoring of operations. Identifying abnormal data patterns and detecting potential failures in these scenarios are important yet rather challenging. In this work, we propose a novel unsupervised anomaly detection method for time series data. The proposed framework jointly learns the observation model and the dynamic model, and model uncertainty is estimated from normal samples. Specifically, a long short-term memory (LSTM)-based encoder-decoder is adopted to represent the mapping between the observation space and the latent space. Bidirectional transitions of states are simultaneously modeled by leveraging backward and forward temporal information. Regularization of the latent space places constraints on the states of normal samples, and Mahalanobis distance is used to evaluate the abnormality level. Empirical studies on synthetic and real-world datasets demonstrate the superior performance of the proposed method in anomaly detection tasks.

Despite the promising results, existing oriented object detection methods usually involve heuristically designed rules, e.g., RRoI generation, rotated NMS. In this paper, we propose an end-to-end framework for oriented object detection, which simplifies the model pipeline and obtains superior performance. Our framework is based on DETR, with the box regression head replaced with a points prediction head. The learning of points is more flexible, and the distribution of points can reflect the angle and size of the target rotated box. We further propose to decouple the query features into classification and regression features, which significantly improves the model precision. Aerial images usually contain thousands of instances. To better balance model precision and efficiency, we propose a novel dynamic query design, which reduces the number of object queries in stacked decoder layers without sacrificing model performance. Finally, we rethink the label assignment strategy of existing DETR-like detectors and propose an effective label re-assignment strategy for improved performance. We name our method D2Q-DETR. Experiments on the largest and challenging DOTA-v1.0 and DOTA-v1.5 datasets show that D2Q-DETR outperforms existing NMS-based and NMS-free oriented object detection methods and achieves the new state-of-the-art., 5 figures

Purpose Overexpression of epithelial cell adhesion molecule (EpCAM) plays essential roles in tumorigenesis and tumor progression in almost all epithelium-derived cancer. Monitoring EpCAM expression in tumors can be used for the diagnosis, staging and prognosis of cancer patients, as well as guiding the individualized treatment of EpCAM-targeted drugs. In this study, we described the synthesis and evaluation of a site-specifically [99mTc]Tc-labeled EpCAM-targeted nanobody for the SPECT/CT imaging of EpCAM expression. Methods We first prepared the [99mTc]Tc-HYNIC-G4K, then it was site-specifically connected to EpCAM-targeted nanobody NB4. The in vitro characteristics of [99mTc]Tc-NB4 were investigated in HT-29 (EpCAM-positive) and HL-60 (EpCAM-negative) cells, while the in vivo studies were performed using small-animal SPECT/CT in the subcutaneous tumor models and the lymph node metastasis model to verify the specific targeting capacity as well as the potential applications of [99mTc]Tc-NB4. Results [99mTc]Tc-NB4 displayed a high EpCAM specificity both in vitro and in vivo. SPECT/CT imaging revealed that [99mTc]Tc-NB4 was cleared rapidly from the blood and normal organs except for the kidneys, and HT-29 tumors were clearly visualized in contrast with HL-60 tumors. The uptake value of [99mTc]Tc-NB4 in HT-29 tumors was increased continuously from 3.77 ± 0.39 %ID/g at 0.5 h to 5.53 ± 0.82 %ID/g at 12 h after injection. Moreover, the [99mTc]Tc-NB4 SPECT/CT could clearly image tumor-infiltrating lymph nodes. Conclusion [99mTc]Tc-NB4 is a broad-spectrum, specific and sensitive SPECT radiotracer for the noninvasive imaging of EpCAM expression in the epithelium-derived cancer, and revealed a great potential for the clinical translation.

It's a meaningful and attractive topic to build a general and inclusive segmentation model that can recognize more categories in various scenarios. A straightforward way is to combine the existing fragmented segmentation datasets and train a multi-dataset network. However, there are two major issues with multi-dataset segmentation: (1) the inconsistent taxonomy demands manual reconciliation to construct a unified taxonomy; (2) the inflexible one-hot common taxonomy causes time-consuming model retraining and defective supervision of unlabeled categories. In this paper, we investigate the multi-dataset segmentation and propose a scalable Language-guided Multi-dataset Segmentation framework, dubbed LMSeg, which supports both semantic and panoptic segmentation. Specifically, we introduce a pre-trained text encoder to map the category names to a text embedding space as a unified taxonomy, instead of using inflexible one-hot label. The model dynamically aligns the segment queries with the category embeddings. Instead of relabeling each dataset with the unified taxonomy, a category-guided decoding module is designed to dynamically guide predictions to each datasets taxonomy. Furthermore, we adopt a dataset-aware augmentation strategy that assigns each dataset a specific image augmentation pipeline, which can suit the properties of images from different datasets. Extensive experiments demonstrate that our method achieves significant improvements on four semantic and three panoptic segmentation datasets, and the ablation study evaluates the effectiveness of each component., 12 figures, 5 figures

IntroductionThe effect of leader pro-social rule breaking on employees is a critical albeit underexplored topic within the domain of study on the consequences of pro-social rule breaking in organizations. This study attempts to make up for the gap by exploring the relationship between leader pro-social rule breaking and employee voice. Drawing on the theory of self-verification, we theorize that leaders who perform pro-social rule breaking will seek feedback from their subordinates, while employees being sought will be triggered to voice upwardly, the extent to which intensity of voice is moderated by the moral courage of employees.MethodsA total of 283 dyads data of supervisor–subordinate from Shanghai, China, in a three-wave time-lagged survey provided support for our hypotheses.ResultsThe results show that leader pro-social rule breaking is positively related to leader feedback-seeking, which is positively related to employee upward voice and mediates the relationship between the two. Moreover, the positive relationship between leader pro-social rule breaking and leader feedback-seeking as well as the indirect effect of leader pro-social rule breaking on employee upward voice via leader feedback-seeking was weakened when moral courage is high.DiscussionThe present study promotes the theoretical research on the positive results of leader pro-social rule breaking and also suggests that feedback-seeking would be an effective way for leaders to motivate employees’ upward voice.

Flowering time influences reproductive success in plants and has a significant impact on yield in grain crops. Flowering time is regulated by a variety of environmental factors, with daylength often playing an important role. Crops can be categorized into different types according to their photoperiod requirements for flowering. For instance, long-day crops include wheat (Triticum aestivum), barley (Hordeum vulgare), and pea (Pisum sativum), while short-day crops include rice (Oryza sativa), soybean (Glycine max), and maize (Zea mays). Understanding the molecular regulation of flowering and genotypic variation therein is important for molecular breeding and crop improvement. This paper reviews the regulation of flowering in different crop species with a particular focus on how photoperiod-related genes facilitate adaptation to local environments.

Existing semantic segmentation works have been mainly focused on designing effective decoders; however, the computational load introduced by the overall structure has long been ignored, which hinders their applications on resource-constrained hardwares. In this paper, we propose a head-free lightweight architecture specifically for semantic segmentation, named Adaptive Frequency Transformer. It adopts a parallel architecture to leverage prototype representations as specific learnable local descriptions which replaces the decoder and preserves the rich image semantics on high-resolution features. Although removing the decoder compresses most of the computation, the accuracy of the parallel structure is still limited by low computational resources. Therefore, we employ heterogeneous operators (CNN and Vision Transformer) for pixel embedding and prototype representations to further save computational costs. Moreover, it is very difficult to linearize the complexity of the vision Transformer from the perspective of spatial domain. Due to the fact that semantic segmentation is very sensitive to frequency information, we construct a lightweight prototype learning block with adaptive frequency filter of complexity $O(n)$ to replace standard self attention with $O(n^{2})$. Extensive experiments on widely adopted datasets demonstrate that our model achieves superior accuracy while retaining only 3M parameters. On the ADE20K dataset, our model achieves 41.8 mIoU and 4.6 GFLOPs, which is 4.4 mIoU higher than Segformer, with 45% less GFLOPs. On the Cityscapes dataset, our model achieves 78.7 mIoU and 34.4 GFLOPs, which is 2.5 mIoU higher than Segformer with 72.5% less GFLOPs. Code is available at https://github.com/dongbo811/AFFormer., Accepted by AAAI2023; codes and models are available at https://github.com/dongbo811/AFFormer

Additional file 1. Fig. S1. SPOCK1and POSTN mainly express in CAF for CRC. (A) Seven major clusters asepithelial, fibroblast, monocyte, endothelial, CMP, B, and T cells inGSE110009. (B) SPOCK1 and POSTN highly express in fibroblast cells inCRC. (C) Seven major clusters as epithelial, macrophage, fibroblast,tissue stem, endothelial, B, and T cellsin GSE120065. (D)SPOCK1 and POSTN highly express in fibroblast cells in CRC.

Additional file 1: Supplementary table 1. The 48 cytokines tested included pro-inflammatory cytokines, chemokines, and growth factors. ABB: abbreviation. Supplementary table 2. Functional ability between the SCD group and NSCD group. Supplementary table 3. Difference of the cytokines level between the NSCD and SCD group at baseline. Supplementary figure 1. Correlation of the cytokines level and PDQ-D scores at baseline. Supplementary figure 2. Relationship between cytokines at baseline and the PDQ-D subtests (A), SDS scores and subtests (B) both at baseline and after 8 weeks treatment.

Parameter-Efficient Transfer Learning (PETL) aims at efficiently adapting large models pre-trained on massive data to downstream tasks with limited task-specific data. In view of the practicality of PETL, previous works focus on tuning a small set of parameters for each downstream task in an end-to-end manner while rarely considering the task distribution shift issue between the pre-training task and the downstream task. This paper proposes a novel two-stage paradigm, where the pre-trained model is first aligned to the target distribution. Then the task-relevant information is leveraged for effective adaptation. Specifically, the first stage narrows the task distribution shift by tuning the scale and shift in the LayerNorm layers. In the second stage, to efficiently learn the task-relevant information, we propose a Taylor expansion-based importance score to identify task-relevant channels for the downstream task and then only tune such a small portion of channels, making the adaptation to be parameter-efficient. Overall, we present a promising new direction for PETL, and the proposed paradigm achieves state-of-the-art performance on the average accuracy of 19 downstream tasks., Comment: 11 pages

Sulpplemental 1 (Fig. S1, Fig. S2 and Fig. S3)

Supplementary information

Human-centric visual tasks have attracted increasing research attention due to their widespread applications. In this paper, we aim to learn a general human representation from massive unlabeled human images which can benefit downstream human-centric tasks to the maximum extent. We call this method SOLIDER, a Semantic cOntrollable seLf-supervIseD lEaRning framework. Unlike the existing self-supervised learning methods, prior knowledge from human images is utilized in SOLIDER to build pseudo semantic labels and import more semantic information into the learned representation. Meanwhile, we note that different downstream tasks always require different ratios of semantic information and appearance information. For example, human parsing requires more semantic information, while person re-identification needs more appearance information for identification purpose. So a single learned representation cannot fit for all requirements. To solve this problem, SOLIDER introduces a conditional network with a semantic controller. After the model is trained, users can send values to the controller to produce representations with different ratios of semantic information, which can fit different needs of downstream tasks. Finally, SOLIDER is verified on six downstream human-centric visual tasks. It outperforms state of the arts and builds new baselines for these tasks. The code is released in https://github.com/tinyvision/SOLIDER., Comment: accepted by CVPR2023

Supplemental document describing the optical setup

Image-text retrieval is a central problem for understanding the semantic relationship between vision and language, and serves as the basis for various visual and language tasks. Most previous works either simply learn coarse-grained representations of the overall image and text, or elaborately establish the correspondence between image regions or pixels and text words. However, the close relations between coarse- and fine-grained representations for each modality are important for image-text retrieval but almost neglected. As a result, such previous works inevitably suffer from low retrieval accuracy or heavy computational cost. In this work, we address image-text retrieval from a novel perspective by combining coarse- and fine-grained representation learning into a unified framework. This framework is consistent with human cognition, as humans simultaneously pay attention to the entire sample and regional elements to understand the semantic content. To this end, a Token-Guided Dual Transformer (TGDT) architecture which consists of two homogeneous branches for image and text modalities, respectively, is proposed for image-text retrieval. The TGDT incorporates both coarse- and fine-grained retrievals into a unified framework and beneficially leverages the advantages of both retrieval approaches. A novel training objective called Consistent Multimodal Contrastive (CMC) loss is proposed accordingly to ensure the intra- and inter-modal semantic consistencies between images and texts in the common embedding space. Equipped with a two-stage inference method based on the mixed global and local cross-modal similarity, the proposed method achieves state-of-the-art retrieval performances with extremely low inference time when compared with representative recent approaches., Comment: Code is publicly available: https://github.com/LCFractal/TGDT

Modern incremental learning for semantic segmentation methods usually learn new categories based on dense annotations. Although achieve promising results, pixel-by-pixel labeling is costly and time-consuming. Weakly incremental learning for semantic segmentation (WILSS) is a novel and attractive task, which aims at learning to segment new classes from cheap and widely available image-level labels. Despite the comparable results, the image-level labels can not provide details to locate each segment, which limits the performance of WILSS. This inspires us to think how to improve and effectively utilize the supervision of new classes given image-level labels while avoiding forgetting old ones. In this work, we propose a novel and data-efficient framework for WILSS, named FMWISS. Specifically, we propose pre-training based co-segmentation to distill the knowledge of complementary foundation models for generating dense pseudo labels. We further optimize the noisy pseudo masks with a teacher-student architecture, where a plug-in teacher is optimized with a proposed dense contrastive loss. Moreover, we introduce memory-based copy-paste augmentation to improve the catastrophic forgetting problem of old classes. Extensive experiments on Pascal VOC and COCO datasets demonstrate the superior performance of our framework, e.g., FMWISS achieves 70.7% and 73.3% in the 15-5 VOC setting, outperforming the state-of-the-art method by 3.4% and 6.1%, respectively., Comment: CVPR 2023

With the progress of 3D human pose and shape estimation, state-of-the-art methods can either be robust to occlusions or obtain pixel-aligned accuracy in non-occlusion cases. However, they cannot obtain robustness and mesh-image alignment at the same time. In this work, we present NIKI (Neural Inverse Kinematics with Invertible Neural Network), which models bi-directional errors to improve the robustness to occlusions and obtain pixel-aligned accuracy. NIKI can learn from both the forward and inverse processes with invertible networks. In the inverse process, the model separates the error from the plausible 3D pose manifold for a robust 3D human pose estimation. In the forward process, we enforce the zero-error boundary conditions to improve the sensitivity to reliable joint positions for better mesh-image alignment. Furthermore, NIKI emulates the analytical inverse kinematics algorithms with the twist-and-swing decomposition for better interpretability. Experiments on standard and occlusion-specific benchmarks demonstrate the effectiveness of NIKI, where we exhibit robust and well-aligned results simultaneously. Code is available at https://github.com/Jeff-sjtu/NIKI, Comment: CVPR 2023

Supplement 1

Meta-learning algorithms are able to learn a new task using previously learned knowledge, but they often require a large number of meta-training tasks which may not be readily available. To address this issue, we propose a method for few-shot learning with fewer tasks, which we call MetaModulation. The key idea is to use a neural network to increase the density of the meta-training tasks by modulating batch normalization parameters during meta-training. Additionally, we modify parameters at various network levels, rather than just a single layer, to increase task diversity. To account for the uncertainty caused by the limited training tasks, we propose a variational MetaModulation where the modulation parameters are treated as latent variables. We also introduce learning variational feature hierarchies by the variational MetaModulation, which modulates features at all layers and can consider task uncertainty and generate more diverse tasks. The ablation studies illustrate the advantages of utilizing a learnable task modulation at different levels and demonstrate the benefit of incorporating probabilistic variants in few-task meta-learning. Our MetaModulation and its variational variants consistently outperform state-of-the-art alternatives on four few-task meta-learning benchmarks., Comment: Accepted by ICML 2023

Additional file 1 Fig. S1. Colon cancer cells promoted the proliferation and migration of Schwann cells by stimulating their secretion of NGF. Fig. S2. Exosomes derived from colon cancer cells facilitated the expression of NGF in Schwann cells via miR-21-5p. Fig. S3 miR-21-5p promoted the expression of NGF in Schwann cells through VHL/HIF-1α. Fig. S4. Schwann cells promoted the proliferation and metastasis of HCT116 cells. Fig. S5. Schwann cells facilitated the proliferation, migration, invasion, and EMT of colon cancer cells through NGF. Fig. S6. NGF facilitated the proliferation, migration, invasion, and EMT of SW480 cells. Fig. S7. NGF facilitated the proliferation, migration, invasion, and EMT of HCT116 cells. Fig. S8. NGF modulated the proliferation and metastasis of colon cancer cells by TrkA. Fig. S9. P75 did not involve in the NGF-induced proliferation and metastasis of colon cancer cells. Fig. S10. P75 did not involve in the NGF-induced proliferation and metastasis of colon cancer cells. Fig. S11. NGF modulated the proliferation and metastasis of colon cancer cells through ERK. Fig. S12. NGF modulated the proliferation and metastasis of colon cancer cells through ERK. Fig. S13. NGF modulated the proliferation and metastasis of colon cancer cells via ERK/ELK1. Fig. S14. NGF modulated the proliferation and metastasis of colon cancer cells via ERK/ELK1. Fig. S15. Schwann cells accelerated the tumorigenesis and metastasis of colon cancer in vivo. Fig. S16. The associated expression of NGF/TrkA/ERK/ELK1/ZEB1/miR-21-5p signaling in colon cancer tissues. Table S1. Clinicopathological characteristics of colon cancer patients. Table S2. Primers of genes in this research for qRT-PCR. Table S3. Details of primary antibodies applied in this study.

Locating 3D objects from a single RGB image via Perspective-n-Point (PnP) is a long-standing problem in computer vision. Driven by end-to-end deep learning, recent studies suggest interpreting PnP as a differentiable layer, allowing for partial learning of 2D-3D point correspondences by backpropagating the gradients of pose loss. Yet, learning the entire correspondences from scratch is highly challenging, particularly for ambiguous pose solutions, where the globally optimal pose is theoretically non-differentiable w.r.t. the points. In this paper, we propose the EPro-PnP, a probabilistic PnP layer for general end-to-end pose estimation, which outputs a distribution of pose with differentiable probability density on the SE(3) manifold. The 2D-3D coordinates and corresponding weights are treated as intermediate variables learned by minimizing the KL divergence between the predicted and target pose distribution. The underlying principle generalizes previous approaches, and resembles the attention mechanism. EPro-PnP can enhance existing correspondence networks, closing the gap between PnP-based method and the task-specific leaders on the LineMOD 6DoF pose estimation benchmark. Furthermore, EPro-PnP helps to explore new possibilities of network design, as we demonstrate a novel deformable correspondence network with the state-of-the-art pose accuracy on the nuScenes 3D object detection benchmark. Our code is available at https://github.com/tjiiv-cprg/EPro-PnP-v2., Comment: Code available at https://github.com/tjiiv-cprg/EPro-PnP-v2. Fixed typos. arXiv admin note: substantial text overlap with arXiv:2203.13254

Model attribution is a critical component of deep neural networks (DNNs) for its interpretability to complex models. Recent studies bring up attention to the security of attribution methods as they are vulnerable to attribution attacks that generate similar images with dramatically different attributions. Existing works have been investigating empirically improving the robustness of DNNs against those attacks; however, none of them explicitly quantifies the actual deviations of attributions. In this work, for the first time, a constrained optimization problem is formulated to derive an upper bound that measures the largest dissimilarity of attributions after the samples are perturbed by any noises within a certain region while the classification results remain the same. Based on the formulation, different practical approaches are introduced to bound the attributions above using Euclidean distance and cosine similarity under both $\ell_2$ and $\ell_\infty$-norm perturbations constraints. The bounds developed by our theoretical study are validated on various datasets and two different types of attacks (PGD attack and IFIA attribution attack). Over 10 million attacks in the experiments indicate that the proposed upper bounds effectively quantify the robustness of models based on the worst-case attribution dissimilarities.

Additional file 2. Fig. S2. Forestmap shows the result of multivariate COX regression analysis for OSamong patients with CRC.

Additional file 3. Fig.S3Correlation of PD-1and TIM-3 with SPOCK1 and POSTN expressions in immunohistologicalstaining of CRCtissues. (A, B) SPOCK1expression is positively correlated with the expressions of PD-1 andTIM-3. (C, D)POSTN expression is positively correlated with the expressions ofPD-1 and TIM-3.Each dotrepresents a sample tissue.

supporting information

Additional file 4. TableS1. Clinicopathological Characteristics of Colorectal Cancer Patients.

Data-driven medium-range weather forecasting has attracted much attention in recent years. However, the forecasting accuracy at high resolution is unsatisfactory currently. Pursuing high-resolution and high-quality weather forecasting, we develop a data-driven model SwinRDM which integrates an improved version of SwinRNN with a diffusion model. SwinRDM performs predictions at 0.25-degree resolution and achieves superior forecasting accuracy to IFS (Integrated Forecast System), the state-of-the-art operational NWP model, on representative atmospheric variables including 500 hPa geopotential (Z500), 850 hPa temperature (T850), 2-m temperature (T2M), and total precipitation (TP), at lead times of up to 5 days. We propose to leverage a two-step strategy to achieve high-resolution predictions at 0.25-degree considering the trade-off between computation memory and forecasting accuracy. Recurrent predictions for future atmospheric fields are firstly performed at 1.40625-degree resolution, and then a diffusion-based super-resolution model is leveraged to recover the high spatial resolution and finer-scale atmospheric details. SwinRDM pushes forward the performance and potential of data-driven models for a large margin towards operational applications.

Additional file 5. Table S2. Patientcharacteristics of CRC from TCGA database.

Deploying reliable deep learning techniques in interdisciplinary applications needs learned models to output accurate and (even more importantly) explainable predictions. Existing approaches typically explicate network outputs in a post-hoc fashion, under an implicit assumption that faithful explanations come from accurate predictions/classifications. We have an opposite claim that explanations boost (or even determine) classification. That is, end-to-end learning of explanation factors to augment discriminative representation extraction could be a more intuitive strategy to inversely assure fine-grained explainability, e.g., in those neuroimaging and neuroscience studies with high-dimensional data containing noisy, redundant, and task-irrelevant information. In this paper, we propose such an explainable geometric deep network dubbed as NeuroExplainer, with applications to uncover altered infant cortical development patterns associated with preterm birth. Given fundamental cortical attributes as network input, our NeuroExplainer adopts a hierarchical attention-decoding framework to learn fine-grained attentions and respective discriminative representations to accurately recognize preterm infants from term-born infants at term-equivalent age. NeuroExplainer learns the hierarchical attention-decoding modules under subject-level weak supervision coupled with targeted regularizers deduced from domain knowledge regarding brain development. These prior-guided constraints implicitly maximizes the explainability metrics (i.e., fidelity, sparsity, and stability) in network training, driving the learned network to output detailed explanations and accurate classifications. Experimental results on the public dHCP benchmark suggest that NeuroExplainer led to quantitatively reliable explanation results that are qualitatively consistent with representative neuroimaging studies., Comment: 12 page 4 fig and 2 table

Localizing people and recognizing their actions from videos is a challenging task towards high-level video understanding. Existing methods are mostly two-stage based, with one stage for person bounding box generation and the other stage for action recognition. However, such two-stage methods are generally with low efficiency. We observe that directly unifying detection and action recognition normally suffers from (i) inferior learning due to different desired properties of context representation for detection and action recognition; (ii) optimization difficulty with insufficient training data. In this work, we present a decoupled one-stage network dubbed DOAD, to mitigate above issues and improve the efficiency for spatio-temporal action detection. To achieve it, we decouple detection and action recognition into two branches. Specifically, one branch focuses on detection representation for actor detection, and the other one for action recognition. For the action branch, we design a transformer-based module (TransPC) to model pairwise relationships between people and context. Different from commonly used vector-based dot product in self-attention, it is built upon a novel matrix-based key and value for Hadamard attention to model person-context information. It not only exploits relationships between person pairs but also takes into account context and relative position information. The results on AVA and UCF101-24 datasets show that our method is competitive with two-stage state-of-the-art methods with significant efficiency improvement., Comment: 11 pages

The quadratic computational complexity to the number of tokens limits the practical applications of Vision Transformers (ViTs). Several works propose to prune redundant tokens to achieve efficient ViTs. However, these methods generally suffer from (i) dramatic accuracy drops, (ii) application difficulty in the local vision transformer, and (iii) non-general-purpose networks for downstream tasks. In this work, we propose a novel Semantic Token ViT (STViT), for efficient global and local vision transformers, which can also be revised to serve as backbone for downstream tasks. The semantic tokens represent cluster centers, and they are initialized by pooling image tokens in space and recovered by attention, which can adaptively represent global or local semantic information. Due to the cluster properties, a few semantic tokens can attain the same effect as vast image tokens, for both global and local vision transformers. For instance, only 16 semantic tokens on DeiT-(Tiny,Small,Base) can achieve the same accuracy with more than 100% inference speed improvement and nearly 60% FLOPs reduction; on Swin-(Tiny,Small,Base), we can employ 16 semantic tokens in each window to further speed it up by around 20% with slight accuracy increase. Besides great success in image classification, we also extend our method to video recognition. In addition, we design a STViT-R(ecover) network to restore the detailed spatial information based on the STViT, making it work for downstream tasks, which is powerless for previous token sparsification methods. Experiments demonstrate that our method can achieve competitive results compared to the original networks in object detection and instance segmentation, with over 30% FLOPs reduction for backbone. Code is available at http://github.com/changsn/STViT-R, Comment: Accepted by CVPR2023

To explore the reasons for the excellent denitrification performance of the Micro-pressure double-cycle reactor (MPDR), the nitrogen removal mechanism of the reactor in the treatment of municipal wastewater was studied. Through analysis of flow simulation and dissolved oxygen (DO) distribution, it was determined that the reactor had a macroscopic biochemical reaction environment for simultaneous nitrification and denitrification (SND) because of the special structure of reactor. The result of sewage treatment showed that the average removal rates of COD, NH4+-N, TN, TP were 92.29%, 96.64%, 73.6% and 91.66% respectively, and the SND rate was 60.9%. Dechloromonas, Thermomonas, Micropruina, Tetrasphaera, etc. for nitrogen and phosphorus removal existed in the reactor at the same time to explain the excellent performance of the system. PICRUSt2 showed that the metabolic pathways related to nutrient degradation in the reactor were highly active and the abundance of denitrification functional genes was higher in the central zone and lower in the peripheral zone. The research results not only perfected the basic theory of the reactor, but more importantly, provided theoretical and technical support for the further application of the reactor.

Efficient exploration of the chemical space to search the candidate drugs that satisfy various constraints is a fundamental task of drug discovery. Advanced deep generative methods attempt to optimize the molecules in the compact latent space instead of the discrete original space, but the mapping between the original and latent spaces is always kept unchanged during the entire optimization process. The unchanged mapping makes those methods challenging to fast adapt to various optimization scenes and leads to the great demand for assessed molecules (samples) to provide optimization direction, which is a considerable expense for drug discovery. To this end, we design a sample-efficient molecular generative method, HelixMO, which explores the scene-sensitive latent space to promote sample efficiency. The scene-sensitive latent space focuses more on modeling the promising molecules by dynamically adjusting the space mapping by leveraging the correlations between the general and scene-specific characteristics during the optimization process. Extensive experiments demonstrate that HelixMO can achieve competitive performance with only a few assessed samples on four molecular optimization scenes. Ablation studies verify the positive impact of the scene-specific latent space, which is capable of identifying the critical characteristics of the promising molecules. We also deployed HelixMO on the website PaddleHelix (https://paddlehelix.baidu.com/app/drug/drugdesign/forecast) to provide drug design service.