Retrieval-augmented generation (RAG) has been proposed to mitigate the hallucination problem of large language models (LLMs) in knowledge-intensive tasks by incorporating external knowledge. However, in multi-hop question answering, existing iterative retrieval methods often struggle to maintain focus on key information. As the number of retrieval iterations increases, the generated queries can gradually drift from the correct reasoning path, and irrelevant or noisy information may accumulate, ultimately reducing reasoning accuracy. To address these challenges, we propose a novel retrieval-augmented generation method for multi-hop question answering based on structured planning. First, our approach employs pre-retrieval question planning to provide semantic guidance for iterative retrieval, ensuring greater consistency between retrieval and reasoning. In addition, we introduce a structured evidence extraction mechanism to effectively filter out noise in the retrieved information, leading to improved reasoning accuracy. Experimental results on three open-domain multi-hop question answering datasets demonstrate that our method can effectively alleviate the impact of retrieval bias and retrieval noise and exhibit competitive performance.

Accurate segmentation of the pancreas and its lesions in CT scans is crucial for the precise diagnosis and treatment of pancreatic cancer. However, it remains a highly challenging task due to several factors such as low tissue contrast with surrounding organs, blurry anatomical boundaries, irregular organ shapes, and the small size of lesions. To tackle these issues, we propose DB-MSMUNet (Dual-Branch Multi-scale Mamba UNet), a novel encoder-decoder architecture designed specifically for robust pancreatic segmentation. The encoder is constructed using a Multi-scale Mamba Module (MSMM), which combines deformable convolutions and multi-scale state space modeling to enhance both global context modeling and local deformation adaptation. The network employs a dual-decoder design: the edge decoder introduces an Edge Enhancement Path (EEP) to explicitly capture boundary cues and refine fuzzy contours, while the area decoder incorporates a Multi-layer Decoder (MLD) to preserve fine-grained details and accurately reconstruct small lesions by leveraging multi-scale deep semantic features. Furthermore, Auxiliary Deep Supervision (ADS) heads are added at multiple scales to both decoders, providing more accurate gradient feedback and further enhancing the discriminative capability of multi-scale features. We conduct extensive experiments on three datasets: the NIH Pancreas dataset, the MSD dataset, and a clinical pancreatic tumor dataset provided by collaborating hospitals. DB-MSMUNet achieves Dice Similarity Coefficients of$89.47 \%, 87.59 \%$, and 89.02 %, respectively, outperforming most existing state-of-the-art methods in terms of segmentation accuracy, edge preservation, and robustness across different datasets. These results demonstrate the effectiveness and generalizability of the proposed method for real-world pancreatic CT segmentation tasks.

Diffusion models have achieved great success in generating 2D images. However, the quality and generaliz-ability of 3D content generation remain limited. State- of-the-art methods often require large-scale 3D assets for training, which are challenging to collect. In this work, we introduce Kiss3DGen (Keep It Simple and Straightforward in 3D Generation), an efficient framework for generating, editing, and enhancing 3D objects by repurposing a well-trained 2D image diffusion model for 3D generation. Specifically, we fine-tune a diffusion model to generate "3D Bundle Image", a tiled representation composed of multi-view images and their corresponding normal maps. The normal maps are then used to reconstruct a 3D mesh, and the multi-view images provide texture mapping, resulting in a complete 3D model. This simple method effectively transforms the 3D generation problem into a 2D image generation task, maximizing the utilization of knowledge in pretrained diffusion models. Furthermore, we demonstrate that our Kiss3DGen model is compatible with various diffusion model techniques, enabling advanced features such as 3D editing, mesh and texture enhancement, etc. Through extensive experiments, we demonstrate the effectiveness of our approach, showcasing its ability to produce high-quality 3D models efficiently Project page: https://ltt-0.github.io/Kiss3dgen.github.io.

Video-to-audio synthesis, which generates synchronized audio for visual content, critically enhances viewer immersion and narrative coherence in film and interactive media.However, video-to-audio dubbing for long-form content remains an unsolved challenge due to dynamic semantic shifts, audio diversity and the absence of dedicated datasets.While existing methods excel in short videos, they falter in long scenarios (e.g., movies) due to fragmented synthesis and inadequate cross-scene consistency.We propose LVAS-Agent, a multi-agent framework that offers a coordinated, multi-component approach to long-video audio generation.Our approach decomposes long-video synthesis into four steps including scene segmentation, script generation, audio design and audio synthesis.To enable systematic evaluation, we introduce LVAS-Bench, the first benchmark with 207 professionally curated long videos spanning diverse scenarios.Experiments show that our method outperforms state-of-the-art V2A models in overall audio synthesis quality.

Due to the low contrast between lesion features and surrounding tissues in non-contrast CT images, traditional detection methods often struggle to accurately identify and differentiate various types of cystic tumors. This limitation increases the risk of misdiagnosis and missed detection, thereby hindering the clinical application of contrast-free techniques. To address this issue, we propose a detection framework that combines the teacher-student model with feature interaction as a non-contrast pancreatic cystic tumor detection technique. The pseudo-labeling of contrast-free CT is detected using the teacher network, which serves as supervisory information for the student network to guide learning. The student network transfers lesion feature information from contrast-enhanced CT to enrich the representation of lesion features. To avoid the confusion of features in multi-phase CT, a multi-domain discriminator is introduced for adversarial learning to extract modality-independent features, which significantly improves the robustness of the model. The experimental results show that the proposed method outperforms the traditional method on plain CT, can effectively detect SCN and MCN lesions, and provides a reliable contrast-free diagnostic solution for the clinic.

We propose PRM, a novel photometric stereo based large reconstruction model to reconstruct high-quality meshes with fine-grained local details. Unlike previous large reconstruction models that prepare images under fixed and simple lighting as both input and supervision, PRM renders photometric stereo images by varying materials and lighting for the purposes, which not only improves the precise local details by providing rich photometric cues but also increases the model robustness to variations in the appearance of input images. To offer enhanced flexibility of images rendering, we incorporate a real-time physically-based rendering (PBR) method and mesh rasterization for online images rendering. Moreover, in employing an explicit mesh as our 3D representation, PRM ensures the application of differentiable PBR, which supports the utilization of multiple photometric supervisions and better models the specular color for high-quality geometry optimization. Our PRM leverages photometric stereo images to achieve high-quality reconstructions with fine-grained local details, even amidst sophisticated image appearances. Extensive experiments demonstrate that PRM significantly outperforms other models.

In this work, we introduce FlexGen, a flexible framework designed to generate controllable and consistent multi-view images, conditioned on a single-view image, or a text prompt, or both. FlexGen tackles the challenges of controllable multi-view synthesis through additional conditioning on 3D-aware text annotations. We utilize the strong reasoning capabilities of GPT-4V to generate 3D-aware text annotations. By analyzing four orthogonal views of an object arranged as tiled multi-view images, GPT-4V can produce text annotations that include 3D-aware information with spatial relationship. By integrating the control signal with proposed adaptive dual-control module, our model can generate multi-view images that correspond to the specified text. FlexGen supports multiple controllable capabilities, allowing users to modify text prompts to generate reasonable and corresponding unseen parts. Additionally, users can influence attributes such as appearance and material properties, including metallic and roughness. Extensive experiments demonstrate that our approach offers enhanced multiple controllability, marking a significant advancement over existing multi-view diffusion models. This work has substantial implications for fields requiring rapid and flexible 3D content creation, including game development, animation, and virtual reality. Project page: https://xxu068.github.io/flexgen.github.io/.

Estimating physical properties for visual data is a crucial task in computer vision, graphics, and robotics, underpinning applications such as augmented reality, physical simulation, and robotic grasping. However, this area remains under-explored due to the inherent ambiguities in physical property estimation. To address these challenges, we introduce GaussianProperty, a training-free framework that assigns physical properties of materials to 3D Gaussians. Specifically, we integrate the segmentation capability of SAM with the recognition capability of GPT-4V(ision) to formulate a global-local physical property reasoning module for 2D images. Then we project the physical properties from multi-view 2D images to 3D Gaussians using a voting strategy. We demonstrate that 3D Gaussians with physical property annotations enable applications in physics-based dynamic simulation and robotic grasping. For physics-based dynamic simulation, we leverage the Material Point Method (MPM) for realistic dynamic simulation. For robot grasping, we develop a grasping force prediction strategy that estimates a safe force range required for object grasping based on the estimated physical properties. Extensive experiments on material segmentation, physics-based dynamic simulation, and robotic grasping validate the effectiveness of our proposed method, highlighting its crucial role in understanding physical properties from visual data. Online demo, code, more cases and annotated datasets are available on \href{https://Gaussian-Property.github.io}{this https URL}.

With the rapid advancement of generative models, general-purpose generation has gained increasing attention as a promising approach to unify diverse tasks across modalities within a single system. Despite this progress, existing open-source frameworks often remain fragile and struggle to support complex real-world applications due to the lack of structured workflow planning and execution-level feedback. To address these limitations, we present ComfyMind, a collaborative AI system designed to enable robust and scalable general-purpose generation, built on the ComfyUI platform. ComfyMind introduces two core innovations: Semantic Workflow Interface (SWI) that abstracts low-level node graphs into callable functional modules described in natural language, enabling high-level composition and reducing structural errors; Search Tree Planning mechanism with localized feedback execution, which models generation as a hierarchical decision process and allows adaptive correction at each stage. Together, these components improve the stability and flexibility of complex generative workflows. We evaluate ComfyMind on three public benchmarks: ComfyBench, GenEval, and Reason-Edit, which span generation, editing, and reasoning tasks. Results show that ComfyMind consistently outperforms existing open-source baselines and achieves performance comparable to GPT-Image-1. ComfyMind paves a promising path for the development of open-source general-purpose generative AI systems.

Abstract Recent large reconstruction models have made notable progress in generating high‐quality 3D objects from single images. However, current reconstruction methods often rely on explicit camera pose estimation or fixed viewpoints, restricting their flexibility and practical applicability. We reformulate 3D reconstruction as image‐to‐image translation and introduce the Relative Coordinate Map (RCM), which aligns multiple unposed images to a “main” view without pose estimation. While RCM simplifies the process, its lack of global 3D supervision can yield noisy outputs. To address this, we propose Relative Coordinate Gaussians (RCG) as an extension to RCM, which treats each pixel's coordinates as a Gaussian center and employs differentiable rasterization for consistent geometry and pose recovery. Our LucidFusion framework handles an arbitrary number of unposed inputs, producing robust 3D reconstructions within seconds and paving the way for more flexible, pose‐free 3D pipelines.

Autonomous surface vessels (ASVs) play a crucial role in military, civilian and scientific research fields, and path planning technology is the foundation for ensuring the navigation of ASVs. However, reinforcement learning algorithms face challenges in path planning, such as low sample utilization efficiency and difficulties in collecting effective data. To address these issues, this paper proposes an efficient reinforcement learning path planning algorithm based on knowledge transfer. First, a state similarity model is introduced to measure the differences between source and target samples. By identifying and transferring similar samples, the learning speed for new tasks is significantly accelerated. Second, considering the complex rules of the International Regulations for Preventing Collisions at Sea (COLREGs), high-quality collision avoidance samples from the source task are selected as knowledge transfer objects, thereby reducing the failure rate of agents in the collision avoidance process. Finally, the impact of the proportion of source task samples in the total sample pool on algorithm performance is analyzed. A progressive exploration method based on sample proportion is established to mitigate the issue of negative transfer. Simulation results demonstrate that the collision avoidance success rate of the algorithm has been significantly improved, with an average success rate of 91% across five typical collision avoidance scenarios—an increase of over 30% compared to the traditional DDPG algorithm. The convergence speed of the reward value improved by 50% after 1000 training episodes. Real-world ship tests verified that it can achieve real-time collision avoidance in complex environments, with a path safety compliance rate of 100%. The algorithm effectively enhances the reliability and safety of autonomous navigation for ASVs.

Vision-centric autonomous driving has recently raised wide attention due to its lower cost. Pretraining is essential for extracting a universal representation. However, current vision-centric pretraining typically relies on either 2D or 3D pre-text tasks, overlooking the temporal characteristics of autonomous driving as a 4D scene understanding task. In this paper, we address this challenge by introducing a world model-based autonomous driving 4D representation learning framework, dubbed DriveWorld, which is capable of pretraining from multi-camera driving videos in a spatiotemporal fashion. Specifically, we propose a Memory State-Space Model for spatiotemporal modelling, which consists of a Dynamic Memory Bank module for learning temporal-aware latent dynamics to predict future changes and a Static Scene Propagation module for learning spatial-aware latent statics to offer comprehensive scene contexts. We additionally introduce a Task Prompt to decouple task-aware features for various downstream tasks. The experiments demonstrate that DriveWorld delivers promising results on various autonomous driving tasks. When pretrained with the OpenScene dataset, DriveWorld achieves a 7.5% increase in mAP for 3D object detection, a 3.0% increase in IoU for online mapping, a 5.0% increase in AMOTA for multi-object tracking, a 0.1m decrease in minADE for motionforecasting, a 3.0% increase in IoU for occupancy prediction, and a 0.34m reduction in average L2 error for planning.

GNN-based social recommendation aims to use social network information to improve recommendation performance of traditional user–item interaction network (U–I network). However, in graph neural network information aggregation, both social networks and U–I networks inevitably have noise, which affects accuracy of recommendation results. To reduce the noise impact of network data, we propose Robust Social Recommendation based on Contrastive Learning and Dual-Stage Graph Neural Network (CLDS). First, considering instability of social networks, we propose the social preference network. It is robust and retains only social friend relationships with common preferences. Based on it and U–I network, we construct a social recommendation pre-training model. Next, we propose self-contrastive learning method. The method initializes multiple social network node representations through Gaussian distribution , pre-training and random disturbance, respectively. Then, it uses contrastive learning on the generated multiple node representations to enhance the robustness of node representation. Finally, CLDS avoids directly capturing potentially user–user and item–item information in U–I networks which is incomplete and untrusted. And instead, it only extracts user–item information to reduce the noise generated by GNN-based U–I network information aggregation. We conduct experiments under the open-source real network dataset. The experimental results show that CLDS outperforms state-of-art methods in social recommendation. The code is available at: https://github.com/Andrewsama/CLDS-master .

Robust 3D perception under corruption has become an essential task for the realm of 3D vision. While current data augmentation techniques usually perform random transformations on all point cloud objects in an offline way and ignore the structure of the samples, resulting in over-or-under enhancement. In this work, we propose an alternative to make sample-adaptive transformations based on the structure of the sample to cope with potential corruption via an auto-augmentation framework, named as Adapt-Point. Specially, we leverage a imitator, consisting of a Deformation Controller and a Mask Controller, respectively in charge of predicting deformation parameters and producing a per-point mask, based on the intrinsic structural information of the input point cloud, and then conduct corruption simulations on top. Then a discriminator is utilized to prevent the generation of excessive corruption that deviates from the original data distribution. In addition, a perception-guidance feedback mechanism is incorporated to guide the generation of samples with appropriate difficulty level. Furthermore, to address the paucity of real-world corrupted point cloud, we also introduce a new dataset ScanObjectNN-C, that exhibits greater similarity to actual data in real-world environments, especially when contrasted with preceding CAD datasets. Experiments show that our method achieves state-of-the-art results on multiple corruption benchmarks, including ModelNet-C, our ScanObjectNN-C, and ShapeNet-C.

3D object detection from the LiDAR point cloud is fundamental to autonomous driving. Large-scale outdoor scenes usually feature significant variance in instance scales, thus requiring features rich in long-range and fine-grained information to support accurate detection. Recent detectors leverage the power of window-based transformers to model long-range dependencies but tend to blur out fine-grained details. To mitigate this gap, we present a novel Mixed-scale Sparse Voxel Transformer, named MsSVT, which can well capture both types of information simultaneously by the divide-and-conquer philosophy. Specifically, MsSVT explicitly divides attention heads into multiple groups, each in charge of attending to information within a particular range. All groups' output is merged to obtain the final mixed-scale features. Moreover, we provide a novel chessboard sampling strategy to reduce the computational complexity of applying a window-based transformer in 3D voxel space. To improve efficiency, we also implement the voxel sampling and gathering operations sparsely with a hash map. Endowed by the powerful capability and high efficiency of modeling mixed-scale information, our single-stage detector built on top of MsSVT surprisingly outperforms state-of-the-art two-stage detectors on Waymo. Our project page: https://github.com/dscdyc/MsSVT.

The multiplicative noise removal problem has received considerable attention recently. To solve this problem, various variational models have been proposed, which minimise an energy functional composed of the data term and the regularisation term. Regarding the regularisation term, a first‐order model is frequently used to remove multiplicative noise, which may cause staircase effect and loss of contrast in the output image. In this study, the authors use a second‐order model, the total curvature (TC), to solve the above problem. The TC model has the benefit of removing the staircase effect and maintaining image edges, contrasts and corners. The augmented Lagrange method is utilised to solve the proposed TC model by introducing auxiliary variables, Lagrange multipliers and using alternating optimisation strategy. In each loop of optimisation, the fast Fourier transform, generalised soft threshold formulas, projection method and gradient descent method are integrated effectively. The experimental results show that the TC model can effectively remove staircase effect and preserve smoothness, via comparison with the first‐order model (total variation regularisation and Perona–Malik regularisation). Furthermore, the TC model is better than another second‐order model based on bounded Hessian regularisation in preserving contrast and corner.

Abstract There are many researches on the polynomial‐based verifiable (k,n) multi‐secret sharing scheme (VMSSS), but none of them focuses on the Chinese remainder theorem (CRT)‐based VMSSS so far. For the first time, we provide a cheater identifiable multi‐secret sharing scheme based on CRT as an alternative method for VMSSS, which is unconditionally secure when the number of cheaters t≤(k − 1)/3. We adopt an encoding method, which makes multiple secrets to be transferred as a single one. In addition, we utilize a single keyed message authenticated code (MAC) to detect and identify cheaters in the reconstruction phase. Then, combine these two methods with a CRT‐based Asmuth‐Bloom's SSS to achieve our design goals. In our scheme, all participants share a single key of MAC rather than each participant possesses an independent key to check the validity of shares, and the size of share is independent in any of n,k, and t. Analyses show that our scheme is more efficient and secure than existing ones. Finally, as an example of the practical impact of our work, we present how our techniques can be applied to secure sum computation. Copyright © 2015 John Wiley & Sons, Ltd.

To compensate for the shortcomings of existing methods used in TSP (Traveling Salesman Problem), such as the accuracy of solutions and the scale of problems, this paper proposed an improved particle swarm optimization by using a self-organizing construction mechanism and dynamic programming algorithm. Particles are connected in way of scale-free fully informed network topology map. Then dynamic programming algorithm is applied to realize the evolution and information exchange of particles. Simulation results show that the proposed method with good stability can effectively reduce the error rate and improve the solution precision while maintaining a low computational complexity.

A new method for performing a nonlinear form of manifold-oriented stochastic neighbor projection method is proposed. By the use of kernel functions, one can operate in the feature space without ever computing the coordinates of the data in that space, but rather by simply computing the inner products between the images of all pairs of data in the feature space. The proposed method is termed as kernel-based manifoldoriented stochastic neighbor projection(KMSNP). By two different strategies, KMSNP is divided into two methods: KMSNP1 and KMSNP2. Experimental results on several databases show that, compared with the relevant methods, the proposed methods obtain higher classification performance and recognition rate. INTRODUCTION Kernel-based methods(kernel methods for short) have become a new hot topic in machine learning fields in recent years, their theoretical basis is statistical learning theory. Kernel methods are a class of algorithms for pattern analysis, whose best known element is the support vector machine(SVM) (Dardas and Georganas 2011).The methods skillfully introduce kernel function which not only reduces the curse of dimensionality (Cherchi and Guevara 2012, Xue et al. 2012), but also effectively solves the local minimum and incomplete statistical analysis in traditional pattern recognition methods on the premise of no additional computational capacity. As an availability way to resolve the problem of nonlinear pattern recognition, kernel methods approach the problem by mapping the data into a highdimensional feature space, where each coordinate corresponds to one feature of the data items, transforming the data into a set of points in a Euclidean space (Chen and Li 2011, Zhang et al. 2008). The theory of kernel methods can be traced back to 1909, Mercer proposed Mercer's theorem (Mercer 1909) which indicates that any ‘reasonable’ kernel function corresponds to some feature space. 1964, the use of Mercer's theorem for interpreting kernels as inner products in a feature space was introduced into machine learning by Aizerman et al. (AizermanI et al. 1964), but no sufficient importance has been attached to it. Until 1992, Vapnik et al. (Boser et al. 1992) successfully extended the SVM to the non-linear SVM by using kernel functions, it began to show its potential and advantages. Subsequently, more and more kernelbased methods were presented, such as: kernel principal component analysis(KPCA) (Xiao et al. 2012), kernel fisher discriminator(KFD) (Yang et al. 2005), kernel independent component analysis (KICA) (Zhang et al. 2013), kernel partial least squares(KPLS) (Helander et al. 2012) and so on. In this paper, we propose to use the kernel idea and present a method called kernel-based manifoldoriented stochastic neighbor projection(KMSNP) method through improving the manifold-oriented stochastic neighbor projection(MSNP) (Wu et al. 2011) technique. MSNP is based on stochastic neighbor embedding(SNE) (Hinton and Roweis 2002) and t-SNE (Maaten and Hinton 2008). The basic principle of SNE is to convert pairwise Euclidean distances into probabilities of selecting neighbors to model pairwise similarities while t-SNE uses student t-distribution to model pairwise dissimilarities in low-dimensional space. Different from SNE and t-SNE, MSNP converts pairwise dissimilarities of inputs to probability distribution related to geodesic distance in highdimensional space and uses Cauchy distribution to model stochastic distribution of features. Furthermore, it recovers the manifold structure through a linear projection by requiring the two distributions to be similar. Experiments demonstrate MSNP has unique advantages in terms of visualization and recognition task, but there are still two drawbacks in it: firstly, MSNP is an unsupervised method and lack of the idea of class label, so it is not suitable for pattern identification; secondly, since MSNP is a linear feature dimensionality reduction algorithm, it cannot effectively settle the nonlinear feature extraction problem. To overcome the disadvantages of MSNP, we have done some preliminary work. On the first, we introduced the idea of class label and presented a method called discriminative stochastic neighbor embedding analysis(DSNE) (Zheng et al. 2012, Chen Proceedings 27th European Conference on Modelling and Simulation ©ECMS Webjorn Rekdalsbakken, Robin T. Bye, Houxiang Zhang (Editors) ISBN: 978-0-9564944-6-7 / ISBN: 978-0-9564944-7-4 (CD) and Wang 2012). On the second, we think KMSNP can overcome the disadvantage mentioned above well. The rest of this paper is organized as follows: in Section 2, we provide a brief review of MSNP. Section 3 describes the detailed algorithm derivation of KMSNP. Furthermore, experiments on various databases are presented in Section 4. Finally, we provide some concluding remarks and describe several issues for future works in Section 5. MSNP Considering the problem of representing d-dimensional data vectors x1, x2, . . . , xN, by r-dimensional (r << d) vectors y1, y2, . . ., yN such that yi represents xi. The basic principle of MSNP is to convert pairwise dissimilarity of inputs to probability distribution related to geodesic distance in high-dimensional space, and then using Cauchy distribution to model stochastic distribution of features, finally, MSNP recovers the manifold structure through a linear projection by requiring the two distributions to be similar. Mathematically, the similarity of datapoint xi to datapoint xj is depicted as the following joint probability pij which means xi how possible to pick xj as its neighbor: exp( / 2) exp( / 2) geo ij ij geo ik k i D p D      (1) where Dij is the geodesic distance for xi and xj. In practice, MSNP calculates geodesic distance by using a two-phase method (Wu et al. 2011). Firstly, an adjacency graph G is constructed by K-nearest neighbor strategy. Secondly, the desired geodesic distance is approximated by the shortest path of graph G. This procedure is proposed in Isomap to estimate geodesic distance and the detail calculation steps can be found in (Tenenbaum et al. 2000). For low-dimensional representations, MSNP employs Cauchy distribution with  degree of freedom to construct joint probability qij. The probability qij indicates how possible point i and point j can be stochastic neighbors is defined as: