Meteorological disasters, especially extreme precipitation, cause significant socioeconomic damage, highlighting the need for effective quantitative precipitation nowcasting. Existing methods, often data-driven and resource-intensive, struggle to capture the underlying physical laws of meteorology. This paper introduces a simple yet effective model using an advection simulator to learn precipitation’s physical dynamics, making the predictions more interpretable. Our model also incorporates a physics-guided module to enhance sensitivity to high-intensity rainfall, improving rainfall prediction accuracy. Experiments on the KNMI radar echo dataset demonstrate that our model outperforms state-of-the-art methods, offering better insights into physics-infused precipitation nowcasting.

Generative modeling of single-cell RNA-seq data is crucial for tasks like trajectory inference, batch effect removal, and simulation of realistic cellular data. However, recent deep generative models simulating synthetic single cells from noise operate on pre-processed continuous gene expression approximations, overlooking the discrete nature of single-cell data, which limits their effectiveness and hinders the incorporation of robust noise models. Additionally, aspects like controllable multi-modal and multi-label generation of cellular data remain underexplored. This work introduces CellFlow for Generation (CFGen), a flow-based conditional generative model that preserves the inherent discreteness of single-cell data. CFGen reliably generates whole-genome, multi-modal, single-cell data, improving the recovery of crucial biological data characteristics while tackling relevant generative tasks such as rare cell type augmentation and batch correction. We also introduce a novel framework for compositional data generation using Flow Matching. By showcasing CFGen on a diverse set of biological datasets and settings, we provide evidence of its value to the fields of computational biology and deep generative models.

Radioguided surgery, such as sentinel lymph node biopsy, relies on the precise localization of radioactive targets by non-imaging gamma/beta detectors. Manual radioactive target detection based on visual display or audible indication of gamma level is highly dependent on the ability of the surgeon to track and interpret the spatial information. This paper presents a learning-based method to realize the autonomous radiotracer detection in robot-assisted surgeries by navigating the probe to the radioactive target. We proposed novel hybrid approach that combines deep reinforcement learning (DRL) with adaptive robotic scanning. The adaptive grid-based scanning could provide initial direction estimation while the DRL-based agent could efficiently navigate to the target utilising historical data. Simulation experiments demonstrate a 95% success rate, and improved efficiency and robustness compared to conventional techniques. Real-world evaluation on the da Vinci Research Kit (dVRK) further confirms the feasibility of the approach, achieving an 80% success rate in radiotracer detection. This method has the potential to enhance consistency, reduce operator dependency, and improve procedural accuracy in radioguided surgeries.

End-to-end paradigm has aroused more and more interests and attention for improving speech-to-text translation (ST) recently. Existing end-to-end models mainly attributes and attempts to address the problem of modeling burden and data scarcity, while always fail to maintain both cross-modal and cross-lingual mapping well at the same time. In this work, we investigate methods for improving endto-end ST with pre-trained acoustic-and-textual models. Our acoustic encoder and decoder begins with processing the source speech sequence as usual. A textual encoder and an adaptor module then obtain source acoustic and textual information respectively, alleviating the representation inconsistency with attentive interactions in the textual decoder. Also, we utilize pre-trained models, and develop an adaptation fine-tuning method to preserve the pre-training knowledge. Experimental results on the IWSLT2023 offline ST task from English to German, Japanese and Chinese show that our method achieves state-of-the-art BLEU scores and surpasses the strong cascaded ST counterparts in unrestricted setting.

This paper presents a 6-DOF hybrid robot for percutaneous needle intervention procedures. The new robot combines the advantages of both serial robots and parallel robots, featuring compactness, high accuracy, and small footprint while overcoming the problems of the high cost of serial robots and the small workspace and singularity issue of parallel robots. Besides, by analyzing the workspace of the robot, the equation is derived between the structure parameter and workspace to adjust the parameters of the robot to satisfy different working scenes. According to the experiment, the accuracy of the robot is related to the position, distance, and insertion angle. The result shows that the performance is better when working near the center workspace and away from the servos and the average error of the robot is 1.39mm. The phantom experiment of lumbar puncture validates its feasibility.

In self-supervised speaker verification, the quality of pseudo labels determines the upper bound of its performance and it is not uncommon to end up with massive amount of unreliable pseudo labels. We observe that the complementary information in different modalities ensures a robust supervisory signal for audio and visual representation learning. This motivates us to propose an audio-visual self-supervised learning framework named Co-Meta Learning. Inspired by the Coteaching+, we design a strategy that allows the information of two modalities to be coordinated through the Update by Disagreement. Moreover, we use the idea of modelagnostic meta learning (MAML) to update the network parameters, which makes the hard samples of two modalities to be better resolved by the other modality through gradient regularization. Compared to the baseline, our proposed method achieves a 29.8%, 11.7% and 12.9% relative improvement on Vox-O, Vox-E and Vox-H trials of Voxceleb1 evaluation dataset respectively.

Automatic speaker verification (ASV) systems are vulnerable to spoofing attacks. As synthetic speech exhibits local and global artifacts compared to natural speech, incorporating local-global dependency would lead to better anti-spoofing performance. To this end, we propose the Rawformer that leverages positional-related local-global dependency for synthetic speech detection. The two-dimensional convolution and Transformer are used in our method to capture local and global dependency, respectively. Specifically, we design a novel positional aggregator that integrates local-global dependency by adding positional information and flattening strategy with less information loss. Furthermore, we propose the squeeze-and-excitation Rawformer (SE-Rawformer), which introduces squeeze-and-excitation operation to acquire local dependency better. The results demonstrate that our proposed SE-Rawformer leads to 37% relative improvement compared to the single state-of-the-art system on ASVspoof 2019 LA and generalizes well on ASVspoof 2021 LA. Especially, using the positional aggregator in the SE-Rawformer brings a 43% improvement on average.

Visual speech (i.e., lip motion) is highly related to auditory speech due to the co-occurrence and synchronization in speech production. This paper investigates this correlation and proposes a cross-modal speech co-learning paradigm. The primary motivation of our cross-modal co-learning method is modeling one modality aided by exploiting knowledge from another modality. Specifically, two cross-modal boosters are introduced based on an audio-visual pseudo-siamese structure to learn the modality-transformed correlation. Inside each booster, a max-feature-map embedded Transformer variant is proposed for modality alignment and enhanced feature generation. The network is co-learned both from scratch and with pretrained models. Experimental results on the test scenarios demonstrate that our proposed method achieves around 60% and 20% average relative performance improvement over baseline unimodal and fusion systems, respectively.

Automatic speaker verification (ASV) suffers from performance degradation in noisy environments. To solve this problem, we propose the noise-disentanglement metric learning to reduce the speaker-irrelevant noisy components and build a noise-invariant embedding space. Specifically, the disentanglement module, including the speaker encoder and re-construction module, is dedicated to decoupling speech signals. The speaker encoder is used to disentangle speaker-related components, and the reconstruction module increases the model’s ability to constrain the noise information by re-constructing the signal. In addition, distribution optimization is introduced to supervise the spatial structure of speaker embeddings under noisy environments. Experiments on Vox-Celeb1 indicate that the proposed method improves the performance of the speaker verification system in both clean and noisy conditions.

Automatic speaker verification (ASV) exhibits unsatisfactory performance under domain mismatch conditions owing to intrinsic and extrinsic factors, such as variations in speaking styles and recording devices encountered in real-world applications. To ensure robust performance under unseen conditions, domain generalization has been explored. However, an inherent contradiction exists between model discrimination and domain generalization, in which the discrimination ability may be reduced while learning to generalize. In this paper, to extract discriminative yet domain-invariant representations, we propose the meta-generalized speaker verification (MGSV) via meta-learning. Specifically, we propose a metric-based distribution optimization and a gradient-based meta-optimization to simultaneously supervise the spatial relationship between embeddings and improve the generalization ability of the model on unseen domains. In addition, we design multiple-single (MS) and simulated speaker verification (SSV) sampling strategies based on single-domain (SD) and single-single (SS) strategies to simulate the train/test domain mismatch more relevantly, thereby mining transferable speaker-related knowledge. SSV is chosen as the most effective method, as it substantially improves the domain generalization by ensuring that the model has learned to discriminate efficiently. Additionally, to intuitively reflect the model performance on the unseen domains, the proposed method is validated on cross-genre, cross-device, and cross-dataset tasks. The experimental results demonstrate that our proposed method achieves remarkable performance in handling domain mismatch issues in speaker verification.

Automatic speaker verification (ASV) faces domain shift caused by the mismatch of intrinsic and extrinsic factors such as recording device and speaking style in real-world applications, which leads to unsatisfactory performance. To this end, we propose the meta generalized transformation via meta-learning to build a domain-invariant embedding space. Specifically, the transformation module is motivated to learn the domain generalization knowledge by executing meta-optimization on the meta-train and meta-test sets which are designed to simulate domain shift. Furthermore, distribution optimization is incorporated to supervise the metric structure of embeddings. In terms of the transformation module, we investigate various instantiations and observe the multilayer perceptron with gating (gMLP) is the most effective given its extrapolation capability. The experimental results on cross-genre and cross-dataset settings demonstrate that the meta generalized transformation dramatically improves the robustness of ASV systems to domain shift, while outperforms the state-of-the-art methods.

Modeling other agents' behaviors plays an important role in decision models for interactions among multiple agents. To optimize its own decisions, a subject agent needs to model what other agents act simultaneously in an uncertain environment. However, modeling insufficiency occurs when the agents are competitive and the subject agent cannot get full knowledge about other agents. Even when the agents are collaborative, they may not share their true behaviors due to their privacy concerns. Most of the recent research still assumes that the agents have common knowledge about their environments and a subject agent has the true behavior of other agents in its mind. Consequently, the resulting techniques are not applicable in many practical problem domains. In this article, we investigate into diversifying behaviors of other agents in the subject agent's decision model before their interactions. The challenges lie in generating and measuring new behaviors of other agents. Starting with prior knowledge about other agents' behaviors, we use a linear reduction technique to extract representative behavioral features from the known behaviors. We subsequently generate their new behaviors by expanding the features and propose two diversity measurements to select top- K $K$ behaviors. We demonstrate the performance of the new techniques in two well-studied problem domains. The top- K $K$ behavior selection embarks the study of unknown behaviors in multiagent decision making and inspires investigation of diversifying agents' behaviors in competitive agent interactions. This study will contribute to intelligent systems dealing with unknown unknowns in an open artificial intelligence world.

The supervisory control and data acquisition (SCADA) system is widely used in industrial control and the contemporary Industrial Internet of Things (IIoT). Unfortunately, due to its relatively weak design in terms of data security and access control, SCADA systems are becoming a favorite target for attackers. End-to-end encryption, such as SSL/TLS protocol, is used to protect the data transmission, but it cannot guarantee security in third-party cloud platforms. In this article, we propose a secure revocable fine-grained access control and data sharing scheme. This scheme not only ensures the confidentiality of the data but also enhances the access control of the SCADA system. Our scheme is based on three key observations. The common communication architecture of SCADA systems cannot protect data security itself. The security supports provided by industrial control protocols are limited. Moreover, the third-party cloud platforms are semitrusted. In addition, we have introduced digital signature technology to assure the integrity of the data in the SCADA system. We prove that our scheme is secure. This scheme has been experimentally evaluated to introduce negligible performance losses while improving data security in the SCADA system.

Audio-visual lip biometrics (AV-LB) has been an emerging biometrics technology that straddles auditory and visual speech processing. Previous works mainly focused on the front-end lip-based feature engineering combined with a shallow statistical back-end model. Over the past decade, convolutional neural network (CNN, or ConvNet) has been widely used and achieved good performance in computer vision and speech processing tasks. However, the lack of a sizeable public AV-LB database led to a stagnation in deep-learning exploration on AV-LB tasks. In addition to the dual audio-visual streams, one essential requirement on the video stream is the region of interest (ROI) around the lips has to be of sufficient resolution. To this end, we compile a moderate-size database using existing public databases. Using this database, we present a deep learning-based AV-LB benchmark, dubbed DeepLip11https://github.com/DanielMengLiu/DeepLip, realized with convolutional video and audio unimodal modules, and a multimodal fusion module. Our experiments show that DeepLip outperforms the traditional lip-biometrics system in context modeling and achieves over 50% relative improvements compared with its unimodal system, with an equal error rate of 0.75% and 1.11% on the test datasets, respectively.

Automatic speaker verification (ASV) has been successfully deployed for identity recognition. With increasing use of ASV technology in real-world applications, channel mismatch caused by the recording devices and environments severely degrade its performance, especially in the case of unseen channels. To this end, we propose a meta speaker embedding network (MSEN) via meta-learning to generate channel-invariant utterance embeddings. Specifically, we optimize the differences between the embeddings of a support set and a query set in order to learn a channel-invariant embedding space for utterances. Furthermore, we incorporate distribution optimization (DO) to stabilize the performance of MSEN. To quantitatively measure the effect of MSEN on unseen channels, we specially design the generalized cross-channel (GCC) evaluation. The experimental results on the HI-MIA corpus demonstrate that the proposed MSEN reduce considerably the impact of channel mismatch, while significantly outperforms other state-of-the-art methods.

Apache Kafka, which is a high-throughput distributed message processing system, has been leveraged by the majority of enterprise for its outstanding performance. Unlike common cloud-based access control architectures, Kafka service providers often need to build their systems on other enterprises' high-performance cloud platforms. However, since the cloud platform belongs to a third party, it is not necessarily reliable. Paradoxically, it has been demonstrated that Kafka's data is stored in the cloud in the plaintext form, and thus poses a serious risk of user privacy leakage. In this paper, we propose a secure fine-grained data transmission scheme called Secure Door on Cloud (SDoC) to protect the data from being leaked in Kafka. SDoC is not only more secure than Kafka's built-in security mechanism, but also can effectively prevent third-party cloud from stealing plaintext data. To evaluate the performance of the SDoC, we simulate normal inter-entity communication and show that Kafka with SDoC integration has a lower data transfer time overhead than that of Kafka with built-in security mechanism opened.

Today's smart devices like fitness tracker, smartwatch, etc., often employ Bluetooth Low Energy (BLE) for data transmission. Such devices thus become our information portal, e.g., SMS message and notifications are delivered to those devices through BLE. In this study, we present BlueDoor, which can obtain unauthorized information from smart devices via BLE vulnerability. We thoroughly examine the BLE protocol, and leverage its intrinsic properties designed for low-cost embedded and wearable devices to bypass the encryption and authentication in BLE. By mimicking a low capacity device to downgrade the process of encryption key negotiation and authentication, BlueDoor can enforce a new key with the peripheral BLE device and pass the authentication without user participation. As a result, BlueDoor can extract BLE packets as well as read/write stored data on BLE devices. We show that BlueDoor works well on the fundamental design tradeoff of using BLE on diverse embedded and wearable devices, and thus can be generalized to various BLE devices. We implement the BlueDoor design and examine its performance on 15 COTS BLE enabled smart devices, including fitness trackers, smartwatch, smart bulb, etc. The results show that BlueDoor can break the information flow and obtain different types of information (e.g., SMS message, notifications) delivered to BLE devices. In addition to privacy threats, this further means traditional operations such as using SMS for verification in widely adopted authentication, are insecure.

With the high popularity of IoT devices, industrial IoT platforms, such as smart factories and oilfield industrial control systems, have become a new trend in the development of smart city. Although various manufacturers pay wide attention to the different functional requirements of IoT platforms, they seldom consider security issues, especially in terms of data security, which has led to a large number of cases of privacy leakage. Some works have been made to provide secure and reliable communication solutions for industrial IoT platforms, unfortunately, as different communication protocols and interaction models are adopted in different scenarios, these solutions are mainly isolated and fragmented. Therefore, it is an urgent challenge to construct a universal cross-platform secure communication scheme for industrial IoT platforms. In this article, we analyze the logic and requirements of different industrial IoT scenarios to abstracts them into a universal model. We summarize the possible attacks on different industrial IoT platforms and design a security scheme to capture these attacks based on the conditional proxy re-encryption primitive. The proposed scheme ensures that data cannot be accessed by an unauthorized user. We also evaluate the security and performance of our scheme, and the experimental results show that our scheme can achieve the functionality and security requirements with low overhead.

This paper studies quantitatively the benefit of startup-failure-driven backup path reprovisioning in terms of connection availability in shared-path-protected optical networks. Numerical results show that shared path protection with startup-failure-driven reprovisioning reduces connection unavailability, and is most efficient when the capacity utilization during the no-failure phases is around 60%.

A class-aware routing scheme is proposed in this paper for OBS networks providing offset-time-based differentiated services (DiffServ). To unravel the structure of the routing problem, mathematical programming formulations are utilized. An optimization model for two-class OBS networks is presented and studied in particular. The objective is to minimize the burst loss probability over the entire network for both traffic classes. The effectiveness of the routing model is evaluated via illustrative examples based on programming techniques. Numerical results show that compared with traditional class-oblivious routing schemes, class-aware routing further reduces the burst loss probability of multiple classes especially when the network load is light or moderate.

This paper investigates backup reprovisioning technique to mitigate the impact of double near-simultaneous link failure scenarios. First, a refined model is introduced for accurate identification of vulnerable connections after the failure of the first link. Then, an improved approach termed Successive Backup Reprovisioning (SBR), is developed for online backup reprovisioning to further reduce the number of affected connections when no extra capacities are added to the network for reprovisioning purpose. Complexity analysis and simulation results show that SBR has advantages in tradeoff between capacity efficiency and execution speed

Traditional optical transport networks are gradually evolving to intelligent optical networks. Intelligent optical networks can support fast bandwidth provision, so they need a faster reconfiguration algorithm. In this paper, we present an algorithm based on simulated annealing for solving the joint logical topology design and routing problem for WDM optical networks. We introduce new mechanism to accelerate the running speed of the simulated annealing algorithm. Computational experiments show that our modified simulated annealing algorithm can obtain a solution very close to that of the classical simulated annealing algorithm, but the running speed of the modified algorithm is much faster.