Fine-Grained Sketch-Based Image Retrieval (FG-SBIR) aims to minimize the distance between sketches and corresponding images in the embedding space. However, scalability is hindered by the growing complexity of solutions, mainly due to the abstract nature of fine-grained sketches. In this paper, we propose an effective approach to narrow the gap between the two domains. It mainly facilitates unified mutual information sharing both intra- and inter-samples, rather than treating them as a single feature alignment problem between modalities. Specifically, our approach includes: (i) Employing dual weight-sharing networks to optimize alignment within the sketch and image domain, which also effectively mitigates model learning saturation issues. (ii) Introducing an objective optimization function based on contrastive loss to enhance the model's ability to align features in both intra- and inter-samples. (iii) Presenting a self-supervised Multi-Scale Token Recycling (MSTR) Module featured by recycling discarded patch tokens in multi-scale features, further enhancing representation capability and retrieval performance. Our framework achieves excellent results on CNN- and ViT-based backbones. Extensive experiments demonstrate its superiority over existing methods. We also introduce Cloths-V1, the first professional fashion sketch-image dataset, utilized to validate our method and will be beneficial for other applications.

In recent years, multi-sensor fusion has gained significant attention from researchers and is used extensively in simultaneous localization and mapping (SLAM) applications, such as visual-inertial odometry (VIO). This technology primarily utilizes visual and odometry measurements for unmanned aerial vehicles (UAVs) to estimate their position, orientation, and environment. However, in most previous works, the input error data of sensors in the system were considered independent. To improve system precision and fully utilize sensor data, a new method called Multi-State Constraint Kalman Filter with NearSAC (MSCKF-NearSAC), based on the MSCKF, is proposed. This method eliminates outliers by limiting the range of selected points, which significantly improves the success rate of feature point matching in the front-end. Furthermore, the MSCKF-ZNN method is proposed for the back-end, and combines zeroing neural network (ZNN) (originated from the Hopfield-type neural network) and error state, resulting in an exponentially converging output trajectory error, thus improving the trajectory precision of the SLAM system. The proposed algorithms, MSCKF-NearSAC and MSCKF-ZNN, are used in the excellent work of the stereo multi-state constraint Kalman filter system (S-MSCKF). A plethora of comparison experiments, utilizing precise measurement and calibration techniques, are conducted on open-source datasets and real-world environments. Experimental results demonstrate that the introduced approach exhibits higher stability in contrast to other algorithms.

Collaboration can amalgamate diverse ideas, styles, and visual elements, fostering creativity and innovation among different designers. In collaborative design, sketches play a pivotal role as a means of expressing design creativity. However, designers often tend to not openly share these meticulously crafted sketches. This phenomenon of data island in the design area hinders its digital transformation under the third wave of AI. In this paper, we introduce a Federated Generative Artificial Intelligence Clothing system, namely StyleWe, employing federated learning to aid in sketch design. StyleWe is committed to establishing an ecosystem wherein designers can exchange sketch styles among themselves. Through StyleWe, designers can generate sketches that incorporate various designers' styles from their peers, drawing inspiration from collaboration without the need for data disclosure or upload. Extensive performance evaluations and user studies indicate that our StyleWe system can produce multi-styled sketches of comparable quality to human-designed ones while significantly enhancing efficiency compared to hand-drawn sketches.

Recently, remote sensing super-resolution (SR) tasks have been widely studied and achieved remarkable performance. However, due to the complex texture and serious image degeneration, the conventional methods (e.g. CNN-based, GAN-based) cannot reconstruct high-resolution (HR) remote sensing images with a large SR factor (≥ ×8). In this paper, we model the large-factor super-resolution (LFSR) task as a referenced diffusion process and explore how to embed pixel-wise constraint into the popular diffusion model. Following this motivation, we propose the first diffusion-based LFSR method named texture consistency diffusion model (TCDM) for remote sensing images. Specifically, we build a novel conditional truncated noise generator (CTNG) in TCDM to simultaneously generate the expectation of posterior probabilityp(xt-1|xt) and the truncated noise image. With the predicted truncated noise image, sampling an SR image using CTNG saves nearly 90% processing time compared to the naive diffusion model. Additionally, we design a new denoising process named texture consistency diffusion (TC-diffusion) to explicitly embed pixel-wise constraints into the LFSR diffusion model during the training stage. Universal experiments on five commonly used remote sensing datasets demonstrate that the proposed TCDM surpasses the latest SR methods by a large margin and reports new SOTA results on several evaluation metrics. Additionally, the proposed method demonstrates impressive visual quality on reconstructed remote sensing image texture and details.

Asynchronous common subset (ACS) is an essential building block for Byzantine fault-tolerance and multi-party computation. The classic ACS framework is due to Ben-Or, Kemler, and Rabin (BKR), consisting of${n}$reliable broadcast (RBC) instances and${n}$asynchronous binary agreement (ABA) instances (where${n}$is the total number of replicas). Despite recent progresses of practical BKR-ACS, the state-of-the-art designs are still trapped by a validity-latency tradeoff. In this paper, we propose Sweeper, a new ACS protocol that breaks the tradeoff, achieving optimal validity and latency. Moreover, Sweeper maintains other benefits including optimal resilience, signature-free, and information-theoretic settings. Sweeper is built on RBC and composable biased reproposable ABA (CBiased RABA). Different from the conventional RABA, CBiased RABA allows replicas to be more biased towards specific RABA instances. We provide generic strategies to transform existing ABA/RABA protocols and a new RABA protocol that we introduce, to CBiased RABA. Furthermore, Sweeper can achieve up to$2 \times $the throughput of PACE-ACS (CCS 2022), the state-of-the-art ACS protocol that follows the BKR-ACS framework.

Quantized neural networks (QNNs) have become a standard operation for efficiently deploying deep learning models on hardware platforms in real application scenarios. An empirical study on German traffic sign recognition benchmark (GTSRB) dataset shows that under the three white-box adversarial attacks of fast gradient sign method, random + fast gradient sign method and basic iterative method, the accuracy of the full quantization model was only 55%, much lower than that of the full precision model (73%). This indicates the adversarial robustness of the full quantization model is much worse than that of the full precision model. To improve the adversarial robustness of the full quantization model, we have designed an adversarial attack defense platform based on field-programmable gate array (FPGA) to jointly optimize the efficiency and robustness of QNNs. Various hardware-friendly techniques such as adversarial training and feature squeezing were studied and transferred to the FPGA platform based on the designed accelerator of QNN. Experiments on the GTSRB dataset show that the adversarial training embedded on FPGA can increase the model's average accuracy by 2.5% on clean data, 15% under white-box attacks, and 4% under black-box attacks, respectively, demonstrating our methodology can improve the robustness of the full quantization model under different adversarial attacks.

Hand-drawn sketches and sketch colourization are the most laborious but necessary steps for fashion designers to design exquisite clothes, especially when the fashion design requires distinctive and personal characteristics from designer style. This paper presents an artificial intelligent aided fashion design system, namely StyleMe, to support the automatic generation of clothing sketches with designer style. Given the clothing pictures specified by the designer, StyleMe can use deep learning based generative model to generate clothing sketches that are consistent with the designer style. The system also supports intelligent colourization on clothing sketch by style transfer, according to specified styles from the real fashion images. Through a series of performance evaluations and user studies, we found that our system can generate effective clothing sketches as good as fashion designers’ human work, and significantly improve the efficiency of fashion design with its sketch colourization method.

Recent years, deep-learning-based methods achieve remarkable improvements on the super-resolution (SR) task. However, recovering high-quality (HQ) texture from the low-quality (LQ) aerial image is still challenging due to the limited contextual modeling ability of current deep-learning methods as well as the sharp artificial texture of aerial images. In this paper, we rethink aerial image super resolution (AISR) task with the perspective of Fourier analysis. Firstly, we build the Fourier Global Convolution (FGC) inspired by the convolution theorem of the Fourier Transform to extract the shadow features. Then, following the Gabor Transform, a carefully designed oriented Texture Contextual Block (OTCB) is proposed to enhance the oriented texture representation. By stacking FGC and OTCB, we propose a simple but effective straight-forward network named Fourier Consistency Image Reconstruction Model (FCIR) to restore HQ aerial image. Moreover, we design a gradient consistency loss (GC Loss) to enhance the quality of reconstructed high-frequency details. Compared with very recent state-of-the-art super-resolution methods, experimental results demonstrate promising SR performance boosts from FCIR on 3 typical aerial image datasets.

Many machine-learning-based intrusion detection methods have been proposed, however there is a lack of collaboration among these methods. Faced with a cascade of malicious behaviors and various running environments, coupled with the endless emergence of new malicious activities, it is difficult for us to choose an algorithm manually that is suitable for all scenarios. In addition, usually the binary detection models are applied that only “normal” or “abnormal” decision is made, and it is difficult for us to know how much confidence we have in the prediction model. In this study, we propose an intrusion collaborative detection framework (ICDF), an ICDF that allows heterogeneous detection models to effectively work together which have complementary expertise. A multialgorithm model ensemble learning method with confidence interval is adopted. In this process, each algorithm model only makes prediction judgments on its own credible probability interval and refuses to predict outside the interval. The final result is generated by voting based on the confidence of multiple models. Ten detection algorithms were tested on three different data sets. Compared with different single algorithms, ICDF could achieve high precision and recall rate, and the best F1 scores.

Deep neural networks have made tremendous progress in 3D object detection, which is an important task especially in autonomous driving scenarios. Benefited from the breakthroughs in deep learning and sensor technologies, 3D object detection methods based on different sensors, such as camera and LiDAR, have developed rapidly. Meanwhile, more and more researches notice that the abundant information contained in the multi-view data can be used to obtain more accurate understanding of the 3D surrounding environment. Therefore, many sensor-fusion 3D object detection methods have been proposed. As safety is critical in autonomous driving and the deep neural networks are known to be vulnerable to adversarial examples with visually imperceptible perturbations, it is significant to investigate adversarial attacks for 3D object detection. Recent works have shown that both image-based and LiDAR-based networks can be attacked by the adversarial examples while the attacks to the sensor-fusion models, which tend to be more robust, haven't been studied. To this end, we propose a simple multi-view correlation based adversarial attack method for the camera-LiDAR fusion 3D object detection models and focus on the black-box attack setting which is more practical in real-world systems. Specifically, we first design a generative network to generate image adversarial examples based on an auxiliary image semantic segmentation network. Then, we develop a cross-view perturbation projection method by exploiting the camera-LiDAR correlations to map each image adversarial example to the space of the point cloud data to form the point cloud adversarial examples in the LiDAR view. Extensive experiments on the KITTI dataset demonstrate the effectiveness of the proposed method.

A distributed less-mass energy control strategy based on SOC of lithium ion battery was designed for aerospace emergency power supply system. Lithium battery SOC measurement is the basis of the control strategy, the article puts forward the project solution to the two problems of the calibration of the battery capacity in the SOC measurement process and the determination of the open circuit voltage and the initial SOC curve. The bidirectional Buck-Boost converter is selected as the energy transmission medium. According to the bus demand, constant voltage discharge mode is chosen. Combined with the characteristics of lithium battery pack, constant current and constant voltage charging method is selected. Based on the above basis, simulation and experimental verification is conducted on the design multiple battery groups coordinated charging and discharging energy management strategies. It also provides a new direction for electric vehicle energy management system.

According to the concept of electromagnetic aeroelastic energy harvesting system, two kinds of magnetic levitation-based electromagnetic energy conversion devices with different winding schemes are designed, and the flutter is simulated by the simple harmonic motion with given amplitude and frequency. The energy conversion capacity of this device was calculated and analyzed. The output voltage and the maximum output power of the device under different amplitude and frequency are compared in two winding schemes, by means of the modeling and analysis of the magnetic field distribution and the transient induced voltage in the induction coil of the electromagnetic energy conversion devices. The simulation results show the effectiveness of the device, which provides a theoretical basis for the design and experiment of the aeroelastic energy harvesting system.