Split learning is a computing resource-friendly distributed learning framework that protects client training data by splitting the model between the client and server. Previous work has proved that split learning faces a severe risk of privacy leakage, as a malicious server can recover the client's private data by hijacking the training process. In this paper, we explore the vulnerability of split learning to server-side backdoor attacks, where our goal is to compromise the model's integrity. Since the server-side attacker cannot access the training data and client model in split learning, the traditional poisoning-based backdoor attack methods are no longer applicable. Therefore, constructing backdoor attacks in split learning poses significant challenges. Our strategy involves the attacker establishing a shadow model on the server side that can encode backdoor samples and guide the client model in learning from this model during the training process, thereby enabling the client to acquire the same capability. Based on these insights, we propose a backdoor attack framework named SFI. Our attack framework minimizes assumptions about the attacker's background knowledge and ensures that the attack remains imperceptible to the client. We implement SFI on various benchmark datasets, and extensive experimental results demonstrate its effectiveness and generality.

Code-based group signatures are a promising candidate for post-quantum cryptography, but existing code-based group signature schemes struggle with the challenges of large signature sizes caused by zero-knowledge proofs. To address this issue, we propose a novel and practical code-based group signature scheme built upon the Schnorr-Lyubashevsky paradigm. Our construction achieves constant-size signatures and public keys, independent of the group cardinality, and its security is formally proven in the random oracle model under the hardness assumptions of the Syndrome Decoding (SD) and Decoding One Out of Many (DOOM) problems. To alleviate the performance bottleneck of rejection sampling, we design and implement a batch processing optimization for the signing algorithm, which significantly accelerates signature generation by applying vectorization to the most computationally intensive operations. Experimental results show that the optimization renders signing practical. Our scheme features the most compact signature size among existing codebased group signature schemes. All related code is open-sourced and available at https://github.com/Latters/CodeBasedGroupSig/.

Split learning has emerged as a practical and efficient privacy-preserving distributed machine learning paradigm. Understanding the privacy risks of split learning is critical for its application in privacy-sensitive scenarios. However, previous attacks against split learning generally depended on unduly strong assumptions or non-standard settings advantageous to the attacker. This paper proposes a novel auxiliary model-based label inference attack framework against learning, namedSplitAUM.SplitAUMfirst builds an auxiliary model on the client side using intermediate representations of the cut layer and a small number of dummy labels. Then, the learning regularization objective is carefully designed to train the auxiliary model and transfer the knowledge of the server model to the client. Finally,SplitAUMuses the auxiliary model output on local data to infer the server’s privacy label. In addition, to further improve the attack effect, we use semi-supervised clustering to initialize the dummy labels of the auxiliary model. SinceSplitAUMrelies only on auxiliary models, it is highly scalable. We conduct extensive experiments on three different categories of datasets, comparing four typical attacks. Experimental results demonstrate thatSplitAUMcan effectively infer privacy labels and outperform existing attack frameworks in challenging yet practical scenarios. We hope our work paves the way for future analyses of the security of split learning.

Split learning is a computing resource-friendly distributed learning framework that protects client training data by splitting the model between the client and server. Previous work has proved that split learning faces a severe risk of privacy leakage, as a malicious server can recover the client's private data by hijacking the training process. In this paper, we first explore the vulnerability of split learning to server-side backdoor attacks, where our goal is to compromise the model's integrity. Since the server-side attacker cannot access the training data and client model in split learning, the traditional poisoning-based backdoor attack methods are no longer applicable. Therefore, constructing backdoor attacks in split learning poses significant challenges. Our strategy involves the attacker establishing a shadow model on the server side that can encode backdoor samples and guiding the client model to learn from this model during the training process, thereby enabling the client to acquire the same capability. Based on these insights, we propose a three-stage backdoor attack framework named SFI. Our attack framework minimizes assumptions about the attacker's background knowledge and ensures that the attack process remains imperceptible to the client. We implement SFI on various benchmark datasets, and extensive experimental results demonstrate its effectiveness and generality. For example, success rates of our attack on MNIST, Fashion, and CIFAR10 datasets all exceed 90%, with limited impact on the main task.

Federated learning (FL) is a privacy-preserving distributed machine learning approach that enables multiple parties to collaboratively train machine learning models without sharing local data. However, compared with the models trained on independent and identically distributed (IID) data, existing methods still face significant degradation in model performance when running on non-IID data. To solve this problem, we propose a federated learning framework based on genetic algorithm and relay strategy in this paper. The framework groups clients according to their local data distribution using genetic algorithm. After obtaining the optimal grouping result, a relay strategy is used to train and aggregate models within each group. Extensive experiments on three benchmark datasets show that FedGR significantly outperforms other state-of-the-art federated learning algorithms on various image classification tasks. The source code is available at https://github.com/zyfhylyh/FedGR.

Leveraging explicit communication and cooperation of multiple robots brings about multiple advantages in the solution of tasks with autonomous robotic agents. For this reason, to the end of transporting polygonal objects with a group of mobile robots, the aim of this article is to develop a fully distributed decision-making and control scheme that lets the robots cooperate as equals, without any kind of central control instance. Apart from coordinating the motions of the robots in a distributed manner, challenges include the self-reliant determination of configurations of the robots around the object, as well as the interrobot negotiation on which robot takes which role or position in the transportation process. Based on some assumptions on the properties of the robots and the considered types of objects, all major aspects of the control scheme are explained in detail, with the most crucial parts making use of optimization-based schemes. This enables unprecedented flexibility by letting the robots automatically adapt to a wide range of scenarios, with different numbers of robots cooperating, and accommodating intricate object shapes. A simulation-based analysis reveals some of the key properties of the proposed scheme, whereas hardware-in-the-loop experiments show the practical applicability with inexpensive hardware.Note to Practitioners—This article was motivated by the potential to further the state of the art of what is possible to achieve with autonomous robots when, instead of using one single robot, a group of robots cooperates self-reliantly to solve a task. The scenario studied in this article is a transportation task in which the robots shall cooperatively push an object to transport it along the desired path. The scheme proposed in this article deals with all major aspects of solving this transportation task, with simulations and hardware-in-the-loop experiments showing the practical applicability of the scheme. While the experiments have been conducted in a laboratory environment, one may readily imagine a more productive application, e.g., in logistics, when employing suitably designed hardware. On another note, maybe more importantly, the general architecture devised to solve the present task may serve as an inspiration or blueprint for the development of decision-making and control schemes that deal with alternative cooperative robotic tasks, especially with regard to the usage of optimization-based schemes on multiple levels allowing a flexible, yet still intuitive problem formulation.

Stigmergy is a common social network mechanism in nature, which allows individuals to achieve effective coordination through the environment, and has shown some successful applications on robotic systems. This paper proposes a stigmergetic target search method based on vector pheromone for swarm robots. Robots generate vector pheromones according to the target signal and release them to the environment for other robots to read, so as to achieve the indirect transfer of information. The method is extended from our previous work which has some shortcoming especially in the updating process of pheromones. This paper improves the previous search method by designing a hybrid pheromone updating rule to achieve more efficient and accurate information exchange. Moreover, the improved search method contains a robot navigation algorithm and a pheromone map building algorithm for guiding robots’ motion and building the pheromone map. Numerical experiments are performed to compare the performance of the improved method with the previous one. Results show that the improved method can enhance the robot’s efficiency and success rate in finding the static and dynamic targets. Meanwhile, it can also improve the speed of map building and the accuracy of the pheromone map.

Deep neural network approaches have made remarkable progress in many machine learning tasks. However, the latest research indicates that they are vulnerable to adversarial perturbations. An adversary can easily mislead the network models by adding well-designed perturbations to the input. The cause of the adversarial examples is unclear. Therefore, it is challenging to build a defense mechanism. In this paper, we propose an image-to-image translation model to defend against adversarial examples. The proposed model is based on a conditional generative adversarial network, which consists of a generator and a discriminator. The generator is used to eliminate adversarial perturbations in the input. The discriminator is used to distinguish generated data from original clean data to improve the training process. In other words, our approach can map the adversarial images to the clean images, which are then fed to the target deep learning model. The defense mechanism is independent of the target model, and the structure of the framework is universal. A series of experiments conducted on MNIST and CIFAR10 show that the proposed method can defend against multiple types of attacks while maintaining good performance.

Swarm robots search for multiple targets in collaboration in unknown environments has been addressed in this paper. An improved grouping strategy based on constriction factors Particle Swarm Optimization is proposed. Robots are grouped under this strategy after several iterations of stochastic movements, which considers the influence range of targets and environmental information they have sensed. The group structure may change dynamically and each group focuses on searching one target. All targets are supposed to be found finally. Obstacle avoidance is considered during the search process. Simulation compared with previous method demonstrates the adaptability, accuracy, and efficiency of the proposed strategy in multiple targets searching.