Object detection with on-board sensors (e.g., lidar, radar, and camera) is crucial to autonomous driving (AD), and these sensors complement each other in modalities. While crowdsensing may potentially exploit these sensors (of huge quantity) to derive more comprehensive knowledge, federated learning (FL) appears to be the necessary tool to reach this potential: it enables autonomous vehicles (AVs) to train machine learning models without explicitly sharing raw sensory data. However, the multimodal sensors introduce various data heterogeneity across distributed AVs (e.g., label quantity skews and varied modalities), posing critical challenges to effective FL. To this end, we present AutoFed as a heterogeneity-aware FL framework to fully exploit multimodal sensory data on AVs and thus enable robust AD. Specifically, we first propose a novel model leveraging pseudo labeling to avoid mistakenly treating unlabeled objects as the background. We also propose an autoencoder-based data imputation method to fill missing data modality (of certain AVs) with the available ones. To further reconcile the heterogeneity, we finally present a client selection mechanism based on client model similarities to improve training stability and convergence rate. Our experiments confirm that AutoFed substantially improves over status quo in both precision and recall, while demonstrating strong robustness to adverse weather.

Channel reassignment is to assign again on the assigned channel resources in order to use the channel resources more efficiently. Channel reassignment in the Software-Defined Networking (SDN) based Internet of Things (SDN-IoT) is a communication promising paradigm, since it allows software-defined routers (SDRs) with the help of SDN controller to appropriately schedule the traffic loads to meet the better transaction of corresponding channels in one link. However, existing channel reassignment works have some limitations: 1) Few works have focused on backbone networks; 2) they have not considered the traffic optimization into the multi-channel reassignment; 3) no one has considered the cooperative multi-channel reassignment between routers. In this paper, we develop a joint multi-channel reassignment and traffic control framework for the core backbone network in SDN-IoT. We develop a Multi-Agent Deep Deterministic Policy Gradient (MADDPG)-based traffic control and multi-channel reassignment (TCCA-MADDPG) algorithm to optimize the objection function to achieve traffic control and channel reassignment. In order to make better use of the time continuity of the channel state, we add an LSTM layer to the neural network in the experiment to capture the timing information of the channel. Simulation results show that the proposed algorithm converges faster and outperform existing methods.

Given the wide adoption of multimodal sensors (e.g., camera, lidar, radar) by autonomous vehicle s (AVs), deep analytics to fuse their outputs for a robust perception become imperative. However, existing fusion methods often make two assumptions rarely holding in practice: i) similar data distributions for all inputs and ii) constant availability for all sensors. Because, for example, lidars have various resolutions and failures of radars may occur, such variability often results in significant performance degradation in fusion. To this end, we present t-READi, an adaptive inference system that accommodates the variability of multimodal sensory data and thus enables robust and efficient perception. t-READi identifies variation-sensitive yet structure-specific model parameters; it then adapts only these parameters while keeping the rest intact. t-READi also leverages a cross-modality contrastive learning method to compensate for the loss from missing modalities. Both functions are implemented to maintain compatibility with existing multimodal deep fusion methods. The extensive experiments evidently demonstrate that compared with the status quo approaches, t-READi not only improves the average inference accuracy by more than 6% but also reduces the inference latency by almost 15× with the cost of only 5% extra memory overhead in the worst case under realistic data and modal variations.

We present Moderator, a policy-based model management system that allows administrators to specify fine-grained content moderation policies and modify the weights of a text-to-image (TTI) model to make it significantly more challenging for users to produce images that violate the policies. In contrast to existing general-purpose model editing techniques, which unlearn concepts without considering the associated contexts, Moderator allows admins to specify what content should be moderated, under which context, how it should be moderated, and why moderation is necessary. Given a set of policies, Moderator first prompts the original model to generate images that need to be moderated, then uses these self-generated images to reverse fine-tune the model to compute task vectors for moderation and finally negates the original model with the task vectors to decrease its performance in generating moderated content. We evaluated Moderator with 14 participants to play the role of admins and found they could quickly learn and author policies to pass unit tests in approximately 2.29 policy iterations. Our experiment with 32 stable diffusion users suggested that Moderator can prevent 65% of users from generating moderated content under 15 attempts and require the remaining users an average of 8.3 times more attempts to generate undesired content.

The volume of data is increasing rapidly currently, which poses a great challenge for resource-constrained clients to process and analyze. A promising approach for solving computation-intensive tasks is to outsource them to the cloud to take advantage of the clouds powerful computing ability. However, it also brings privacy and security issues since the data uploaded to the cloud may contain sensitive and private information which should be protected. In this paper, we address this problem and focus on the privacy-preserving and secure outsourcing of large-scale nonlinear programming problems (NLPs) subject to both linear constraints and nonlinear constraints. Large-scale NLPs play an important role in the eld of data analytics but have not received enough attention. In our outsourcing protocol, we rst apply a secure and efcient transformation scheme at the client side to encrypt its private information. Then, we use the reduced gradient method and generalized gradient method at the server side to solve the transformed large-scale NLPs under linear constraints and nonlinear constraints, respectively. We provide security analysis of the proposed protocol, and evaluate its performance. The experimental results show that we can efciently solve the large-scale NLPs and save huge amount of time for the clients.