Search

Deep neural network (DNN) -based object detectors have achieved remarkable success, but recent research has revealed their vulnerability to backdoor attacks. The attacks cause the poisoned model to output results normally on benign images, but outputs results specified by the attacker on images inserted with a trigger. Although backdoor attacks have been extensively investigated on image classification tasks, their exploration in object detection tasks remains limited. With the increasing application of object detectors in safety-sensitive fields such as autonomous driving, backdoor attacks on object detection tasks may have serious consequences. Currently, strategies for object disappearance attack scenarios exhibit certain limitations. First, these strategies typically exhibit a one-to-one correspondence, implying that the insertion of one trigger can only result in the disappearance of one object. Second, these strategies typically necessitate the attacker’s knowledge of the object’s precise location information to achieve its disappearance, thereby rendering real-time insertion of triggers unfeasible. Finally, these strategies exhibit diminished attack success rates when applied to two-stage detectors. The paper presents a global object disappearance attack scenario and proposes a simple, covert, and highly effective attack strategy. Experimental evaluations are conducted on four widely-used object detection models (Yolov5s, Yolov8s, Faster R-CNN, and Libra R-CNN) using two benchmark datasets (PASCAL VOC $07+12$ and MS COCO2017) to validate the effectiveness of the proposed strategy. The results demonstrate that the success rate of this attack strategy exceeds 96%, while the poison rate is only 10%.

Polypoid lesions in the uterus are very common. There are a lot of benign lesions and malignant tumor should be considered. Adenosarcoma, one of the differential diagnoses, is a rare mixed epithelial and mesenchymal tumor consisting of benign epithelial components and sarcoma stroma. Here, we present a case of atypical adenosarcoma that has never been reported. This tumor was composed of benign endometrial epithelium and hyperplastic stroma with extensive bizarre stromal cells. But the cleft-like spaces and papillary stomal fronds which were the typical histological images of adenosarcoma were absent. These stomal cells, including bizarre cells, were positive for vimentin, CD10, ER, PR, cyclin D1 and P16 but were immunonegative for caldesmon. Furthermore, this tumor harbored amplification of MDM2, as revealed by fluorescence in situ hybridization testing (FISH) and next-generation DNA sequencing (NGS).

Federated learning framework facilitates more applications of deep learning algorithms on the existing network architectures, where the model parameters are aggregated in a centralized manner. However, some of federated learning participants are often inaccessible, such as in a power shortage or dormant state. That will force us to explore the possibility that the parameter aggregation is operated in an ad hoc manner, which is based on consensus computing. On the contrary, since caching mechanism is indispensable to any federated learning mobile node, it is necessary to investigate the connection between it and consensus computing. In this article, we first propose a novel federated learning paradigm, which supports an ad hoc operation mode for federated learning participants. Second, a discrete-time dynamic equation and its control law are formulated to satisfy the demands from federated learning framework, with a quantized caching scheme designed to mask the uncertainties from both asynchronous updates and measurement noises. Then, the consensus conditions and the convergence of the consensus protocol are deduced analytically, and a quantized caching strategy to optimize the convergence speed is provided. Our major contribution is to give the basic theories of distributed consensus problem for federated learning framework, and the theoretical results are validated by numerical simulations.

Because images are vulnerable to external attacks in the process of network transmission and traditional image encryption algorithms have limitations such as long encryption time, insufficient entropy or poor diffusion of cipher image information when encrypting color images, a fast image encryption algorithm based on logistics-sine-cosine mapping is proposed. The algorithm first generates five sets of encrypted sequences from the logistics-sine-cosine mapping, then uses the order of the encryption sequence to scramble the image pixels and designs a new pixel diffusion network to further improve the key sensitivity and plain-image sensitivity of the encryption algorithm. Finally, in a series of security analysis experiments, the experimental image Lena was tested 100 times, and the average encryption time was 0.479 s. The average value of the information entropy, pixel change rate and uniform average change intensity of the cipher image reached 7.9994, 99.62% and 33.48%, respectively. The experimental results show that the fast image encryption algorithm based on logistics-sine-cosine mapping takes less time to encrypt, and the cipher image has good information entropy and diffusivity. It is a safe and effective fast image encryption algorithm.