High-precision underwater navigation is a prerequisite for underwater vehicles to explore the marine environment and resources. Inertial navigation system (INS), Doppler velocity logger (DVL) and Pressure sensor (PS) onboard the underwater vehicles are favored for integrate due to their effectiveness and user-friendly configuration. Specifically, the velocity measurement from the DVL is important to reduce the accumulated error of the INS. Different from the traditional filter-based fusion solution, we firstly introduce a Factor graph optimization (FGO) framework to integrate the INS/DVL/PS navigation system. Since the DVL observations often experience outliers and outage due to the harsh sea condition and range limitation, it can lead to rapid position drift and error accumulation in the navigation system. Thus, we develop an outlier detection scheme based on the improved interquartile range method, which combines sliding window and dynamic threshold adjustment of the propeller revolutions per minute (RPM). Then, we propose a DVL velocity prediction method based on the nonlinear least squares (NLS)-Transformer-LSTM model. We refine the RPM by NLS and use it as one of the key features in the prediction model. Finally, the constructed system is compared with several classical Kalman filters through both simulated and measured experiments. The performance of the NLS-Transformer-LSTM model is comprehensively evaluated. The results indicate that the system can provide more accurate pseudo-DVL velocity estimates. This capability ensures more stable and reliable underwater navigation accuracy during DVL outages. Consequently, it offers strong support for autonomous underwater vehicle applications in complex marine environments.

Existing modification-based linguistic steganography methods primarily perform linguistic manipulations within a single embedding space to conceal secret information. However, these methods are stringently constrained by the original semantics of the cover text, making it struggle to achieve a satisfactory embedding capacity in a single embedding space. In this paper, we propose a novel Multi-granularity Modification-based Linguistic Steganography framework (MMLS) that hides secret information in both syntactic space and symbolic space, enhancing syntactic naturalness and semantic coherence while further increasing embedding capacity. Specifically, MMLS utilizes a paraphrase generation model to automatically modify the syntactic structure of the given original sentence, which enables the generation of paraphrases and the preservation of semantics simultaneously. Moreover, MMLS employs a distance-aware syntactic bins coding strategy to embed part of secret information into the syntactic space. This strategy utilizes a cluster-based way to partition the implicit syntactic space into a finite number of separate zones, thus increasing the number of candidate paraphrases and avoiding the selection of semantically distorted steganographic texts. Finally, the pre-trained BERT is used to replace some words in candidate paraphrases with their synonyms. Such a design embeds the remaining secret information into symbolic space while ensuring syntactic and semantic naturalness. Experimental results demonstrate that MMLS significantly outperforms existing methods in terms of semantic coherence, embedding capacity, and security.

One major challenge in weakly supervised learning is learning from inexact supervision, ranging from partial labels (PLs) with redundant information to the extreme of unlabeled data with insufficient information. While recent work has made significant strides in specific inexact supervision contexts, supervision forms typically coexist in complex combinations. This is exemplified in semi-supervised partial label learning, where PLs act as the exclusive supervision in a semi-supervised setting. Current strategies addressing combined inexact scenarios are usually composite, which can lead to incremental solutions that essentially replicate existing methods. In this paper, we propose a novel approach to uniformly tackle both label redundancy and insufficiency, derived from a mutual information-based perspective. We design a label channel that facilitates dynamic label exchange within the candidate label sets, which identifies potential true labels and filters out likely incorrect ones, thereby minimizing error accumulation. Experimental results demonstrate the superiority of our method over existing state-of-the-art PL and semi-supervised learning approaches by directly integrating them. Furthermore, our extended experiments on partial-complementary label learning underscore the flexibility of our uniform treatment in managing diverse supervision scenarios.

As the architectural complexity of processors increases dramatically, rigorous functional verification remains essential to ensure performance and immunity from design bugs. A pivotal yet challenging aspect of functional verification is the generation of test instruction streams that are not only highly effective in coverage but also compact enough to significantly reduce verification time. In this paper, we introduce DIG, a novel Dynamic Instruction Generator that leverages processor runtime architectural states through an instruction set simulator. In essence, DIG accesses processor runtime information to produce instruction streams with valid semantics, effectively avoiding illegal memory accesses and infinite loops. The quality of these instruction streams is further enhanced by incorporating both intra-instruction and inter-instruction test knowledge. The effectiveness of DIG is demonstrated through a case study involving Western Digital’s open-source RISC-V core, VeeR EH2. Our experimental results indicate that DIG reduces the number of test instructions by as much as 62.50% and 86.11%, compared to the state-of-the-art random instruction generators, RISC-V DV and RISC-V Torture, respectively, while simultaneously achieving superior functional coverage.

A partial label (PL) specifies a set of candidate labels for an instance and partial-label learning (PLL) trains multi-class classifiers with PLs. Recently, many methods that incorporate techniques from other domains have shown strong potential. The expectation that stronger techniques would enhance performance has resulted in prominent PLL methods becoming not only highly complicated but also quite different from one another, making it challenging to choose the best direction for future algorithm design. While it is exciting to see higher performance, this leaves open a fundamental question: what makes a PLL method effective? We present a comprehensive empirical analysis of this question and summarize the success of PLL so far into some minimal algorithm design principles. Our findings reveal that high accuracy on benchmark-simulated datasets with PLs can misleadingly amplify the perceived effectiveness of some general techniques, which may improve representation learning but have limited impact on addressing the inherent challenges of PLs. We further identify the common behavior among successful PLL methods as a progressive transition from uniform to one-hot pseudo-labels, highlighting the critical role of mini-batch PL purification in achieving top performance. Based on our findings, we introduce a minimal working algorithm that is surprisingly simple yet effective, and propose an improved strategy to implement the design principles, suggesting a promising direction for improvements in PLL.

Vision transformers (ViTs) have become one of the dominant frameworks for vision tasks in recent years because of their ability to efficiently capture long-range dependencies in image recognition tasks using self-attention. In fact, both CNNs and ViTs have advantages and disadvantages in vision tasks, and some studies suggest that the use of both may be an effective way to balance performance and computational cost. In this paper, we propose a new hybrid network based on CNN and transformer, using CNN to extract local features and transformer to capture long-distance dependencies. We also proposed a new feature map resolution reduction based on Discrete Cosine Transform and self-attention, named DCT-Attention Down-sample (DAD). Our DctViT-L achieves 84.8% top-1 accuracy on ImageNet 1K, far outperforming CMT, Next-ViT, SpectFormer and other state-of-the-art models, with lower computational costs. Using DctViT-B as the backbone, RetinaNet can achieve 46.8% mAP on COCO val2017, which improves mAP by 2.5% and 1.1% with less calculation cost compared with CMT-S and SpectFormer as the backbone.

SARS-CoV-2 transmission in indoor spaces, where most infection events occur, depends on the types and duration of human interactions, among others. Understanding how these human behaviours interface with virus characteristics to drive pathogen transmission and dictate the outcomes of non-pharmaceutical interventions is important for the informed and safe use of indoor spaces. To better understand these complex interactions, we developed the Pedestrian Dynamics-Virus Spread model (PeDViS), an individual-based model that combines pedestrian behaviour models with virus spread models incorporating direct and indirect transmission routes. We explored the relationships between virus exposure and the duration, distance, respiratory behaviour, and environment in which interactions between infected and uninfected individuals took place and compared this to benchmark 'at risk' interactions (1.5 metres for 15 minutes). When considering aerosol transmission, individuals adhering to distancing measures may be at risk due to the buildup of airborne virus in the environment when infected individuals spend prolonged time indoors. In our restaurant case, guests seated at tables near infected individuals were at limited risk of infection but could, particularly in poorly ventilated places, experience risks that surpass that of benchmark interactions. Combining interventions that target different transmission routes can aid in accumulating impact, for instance by combining ventilation with face masks. The impact of such combined interventions depends on the relative importance of transmission routes, which is hard to disentangle and highly context dependent. This uncertainty should be considered when assessing transmission risks upon different types of human interactions in indoor spaces. We illustrated the multi-dimensionality of indoor SARS-CoV-2 transmission that emerges from the interplay of human behaviour and the spread of respiratory viruses. A modelling strategy that incorporates this in risk assessments can help inform policy makers and citizens on the safe use of indoor spaces with varying inter-human interactions.

The autoregressive-based model with the left-to-right generation order has been a predominant paradigm for generative linguistic steganography. However, such steganography does not perform well on semantic control and content planning, which is forced by the secret message during the generation process. To mitigate this issue and efficiently produce high-quality steganographic texts (stegotexts), we present aProgressiveNon-autoregressiveGenerative linguisticSteganography (PNG-Stega), which encodes secret messages and extends the context to generate stegotexts in a multi-round insertion manner. Each round continuously refines the generated steganographic sequences on the premise of the global information of the previous round, while striving to decline the adverse effects of steganographic encoding on text quality. Moreover, for enhancing the semantic internal dependency of stegotexts, we utilize a constraint word sequences extraction scheme to obtain keywords to initialize the skeleton of targeted stegotexts, then expand the existing keywords with insertion operations. Experimental results demonstrate that PNG-Stega outperforms compared methods in terms of imperceptibility and anti-steganalysis ability. In particular, PNG-Stega provides high information hiding efficiency, even exceeding the autoregressive methods by around 2 times.

The current lexical substitution-based linguistic steganography primarily determines the substitutions by their linguistic suitability, overlooking the disparity in their capability against steganalysis. This oversight leads to the potential security risk by performing poor substitutions. To address this issue, this letter proposes aCausalPerceptionGuidedLinguisticSteganography(CPG-LS) via elaborate and secure lexical substitutions. CPG-LS constructs a causal perception network by making full use of a trained CNN discriminator to assess the security of each original word in the cover text and its substitutable candidates for controlling the embedding of secret message. Particularly, the causal score of each word is comprehensively measured from two perspectives by the causal perception network, one is word saliency, and the other is anti-steganalysis capability. Finally, the selection of original words is explicitly guided by their causal scores for securely embedding secret message. As the causal score of a word decreases, the perturbation caused by modifying this word becomes smaller, leading to stronger anti-steganalysis capability and lower semantic distortion in the obtained stegotexts. The experimental results demonstrate that CPG-LS achieves better text quality and anti-steganalysis capability than existing similar methods.

Precipitable water vapor (PWV) is an important parameter reflecting atmospheric water vapor, which plays an important role in the global hydrological cycle. At present, brightness temperatures (Tbs) at frequencies of 18 and 23 GHz from the Advanced Microwave Scanning Radiometer 2 (AMSR2) are commonly applied in PWV data retrieval. Two tools are usually employed to perform the retrieval: physical models and neural networks. In this article, we creatively deduced a more universal physical equation and verified the consistency between theoretical and experimental equations both theoretically and numerically. Then, we fully considered the defects of these two tools and proposed a new integrated method that combines the physical model with a neural network. The Tbs in horizontal and vertical polarizations from AMSR2 and GNSS-derived PWV (GNSS-PWV) data from the SuomiNet global network were used to build the physical model. In the construction of the neural network, we also introduced the related surface parameters from GNSS stations as inputs. Validation results obtained with GNSS-PWV data showed that the accuracy of the test set can reach 2.38 and 2.37 mm in the ascending and descending orbit cases, respectively. Compared with the traditional physical model and neural network model, the improvements of the test set were 24.4% and 17.4% in the ascending orbit and 26.4% and 19.4% in the descending orbit. Radiosonde observation (RAOB) data were applied to carry out another external independent verification, and the accuracy of the test set reached 2.70 and 3.54 mm based on the RAOB data.