Fusing cross-category objects to a single coherent object has gained increasing attention in text-to-image (T2I) generation due to its broad applications in virtual reality, digital media, film, and gaming. However, existing methods often produce biased, visually chaotic, or semantically inconsistent results due to overlapping artifacts and poor integration. Moreover, progress in this field has been limited by the absence of a comprehensive benchmark dataset. To address these problems, we propose Adaptive Group Swapping (AGSwap), a simple yet highly effective approach comprising two key components: (1) Group-wise Embedding Swapping, which fuses semantic attributes from different concepts through feature manipulation, and (2) Adaptive Group Updating, a dynamic optimization mechanism guided by a balance evaluation score to ensure coherent synthesis. Additionally, we introduce Cross-category Object Fusion (COF), a large-scale, hierarchically structured dataset built upon ImageNet-1K and WordNet. COF includes 95 superclasses, each with 10 subclasses, enabling 451,250 unique fusion pairs. Extensive experiments demonstrate that AGSwap outperforms state-of-the-art compositional T2I methods, including GPT-Image-1 using simple and complex prompts. Project Page

In this paper, we study an object synthesis task that combines an object text with an object image to create a new object image. However, most diffusion models struggle with this task, \textit{i.e.}, often generating an object that predominantly reflects either the text or the image due to an imbalance between their inputs. To address this issue, we propose a simple yet effective method called Adaptive Text-Image Harmony (ATIH) to generate novel and surprising objects. First, we introduce a scale factor and an injection step to balance text and image features in cross-attention and to preserve image information in self-attention during the text-image inversion diffusion process, respectively. Second, to better integrate object text and image, we design a balanced loss function with a noise parameter, ensuring both optimal editability and fidelity of the object image. Third, to adaptively adjust these parameters, we present a novel similarity score function that not only maximizes the similarities between the generated object image and the input text/image but also balances these similarities to harmonize text and image integration. Extensive experiments demonstrate the effectiveness of our approach, showcasing remarkable object creations such as colobus-glass jar. https://xzr52.github.io/ATIH/

Robot-assisted gastrointestinal endoscopic surgery (GES) as a kind of natural orifice transluminal endoscopic surgery (NOTES) is the next-generation minimally invasive surgery (MIS). Besides, rendering certain autonomy to a Gas-trointestinal Endoscopic Surgical Robot (GESR) is promising but highly challenging. Therefore, to accelerate the development and augment the autonomy of GESR, we use CoppeliaSim to develop the first robotic simulator for the GESR system (GESRsim) based on our previous design. The GESRsim provides several 3D models and kinematics of our designed manipulators and endoscopic snake bone. Additionally, we build several scenes for robotic GES training and then utilize different programming interfaces to perform teleoperation. Furthermore, several advanced control algorithms, including visual servoing (VS) and deep reinforcement learning (DRL), are implemented to verify the performance of the GESRsim.

Although we may observe heterogeneous traffic appearance on the network backbone, malicious traffic tends to converge with their traffic appearance similarity due to the consistent hostile behaviors of the same anomaly category. Measuring such traffic similarity of host behaviors can help us to detect anomalous traffic from benign traffic. This paper proposes a novel framework for the detection of network intrusion based on traffic similarity measures and clustering. We apply the grouping and DBSCAN method to feature dimensionality reduction so that traffic carrying the same category anomalies is concentrated in the limited amount of clusters, which can be interpreted as the structured significant characteristics of the corresponding anomaly category. The derived anomaly cluster characteristics are useful for detecting newly coming traffic in future for its maliciousness. Based on the experiment with the IDS 2018 dataset, our proposed detection procedure can effectively separate the malicious network traffic from background with an accuracy of up to 96%. Our proposed method has apparent benefits for identifying malicious traffic in large-scale network traffic data, and it is a practical intrusion detection method.

In response to the surging challenge in the number and types of mobile malware targeting smart devices and their sophistication in malicious behavior camouflage, we propose to compose a traffic behavior modeling method based on one-dimensional convolutional neural network with autoencoder and independent recurrent neural network (1DCAE-IndRNN) for mobile malware detection. The design solves the problem that most existing approaches for mobile malware traffic detection struggle with capturing the network traffic dynamics and the sequential characteristics of anomalies in the traffic. We reconstruct and apply the one-dimensional convolutional neural network to extract local features from multiple network flows. The autoencoder is applied to digest the principal traffic features from the neural network and is integrated into the independent recurrent neural network construction to highlight the sequential relationship between the highly significant features. In addition, the Softmax function with the LReLU activation function is adjusted and embedded to the neurons of the independent recurrent neural network to effectively alleviate the problem of unstable training. We conduct a series of experiments to evaluate the effectiveness of the proposed method and its performance for the 1DCAE-IndRNN-integrated detection procedure. The detection results of the public Android malware dataset CICAndMal2017 show that the proposed method achieves up to 98% detection accuracy and recall rates with clear advantages over other benchmark methods.