The Diffusion Transformer plays a pivotal role in advancing text-to-image and text-to-video generation, owing primarily to its inherent scalability. However, existing controlled diffusion transformer methods incur significant parameter and computational overheads and suffer from inefficient resource allocation due to their failure to account for the varying relevance of control information across different transformer layers. To address this, we propose the Relevance-Guided Efficient Controllable Generation framework, RelaCtrl, enabling efficient and resource-optimized integration of control signals into the Diffusion Transformer. First, we evaluate the relevance of each layer in the Diffusion Transformer to the control information by assessing the ControlNet Relevance Score, which measures the impact of skipping each control layer on both the quality of generation and the control effectiveness during inference. Based on the strength of the relevance, we then tailor the positioning, parameter scale, and modeling capacity of the control layers to reduce unnecessary parameters and redundant computations. Additionally, to further improve efficiency, we replace the self-attention and FFN in the commonly used copy block with the carefully designed Two-Dimensional Shuffle Mixer (TDSM), enabling efficient implementation of both the token mixer and channel mixer. Both qualitative and quantitative experimental results demonstrate that our approach achieves superior performance with only 15% of the parameters and computational complexity compared to PixArt-delta.

State space models (SSMs) have recently garnered significant attention in computer vision. However, due to the unique characteristics of image data, adapting SSMs from natural language processing to computer vision has not outperformed the state-of-the-art convolutional neural networks (CNNs) and Vision Transformers (ViTs). Existing vision SSMs primarily leverage manually designed scans to flatten image patches into sequences locally or globally. This approach disrupts the original semantic spatial adjacency of the image and lacks flexibility, making it difficult to capture complex image structures. To address this limitation, we propose Dynamic Adaptive Scan (DAS), a data-driven method that adaptively allocates scanning orders and regions. This enables more flexible modeling capabilities while maintaining linear computational complexity and global modeling capacity. Based on DAS, we further propose the vision backbone DAMamba, which significantly outperforms popular vision Mamba models in vision tasks such as image classification, object detection, instance segmentation, and semantic segmentation. Notably, it surpasses some of the latest state-of-the-art CNNs and ViTs.

The Industrial Internet of Things (IIoT) enables communication among automation systems, machinery, and sensors in an industrial setting. To optimize critical industrial operations, a substantial volume of data concerning diverse in-factory activities and automation services is generated by IoT devices and sensors. This data are subsequently transferred to distant processing systems for analysis and decision-making. Nevertheless, a substantial latency in data transmission or any abnormality in the generated data may result in delayed or erroneous decisions, consequently impacting the efficacy of essential industrial systems. To address these challenges, we established an intelligent network architecture utilizing software-defined networking that achieves tactile latencies efficiently while handling industrial data traffic in an energy-efficient manner. To address the initial challenge, the suggested architecture utilizes the self-organized maps approach to distinguish between industrial traffic requiring tactile latencies and nontactile traffic. We utilize a binary tree-based flow table mapping method to enhance flow table matching and decrease lookup times. To address the second challenge, we employ the Support Vector Machine technique to identify anomalies in real-time industrial data traffic. The Hadoop system and Mininet emulator are utilized to evaluate the proposed architecture using the UNSW dataset. The results demonstrate the effectiveness of the suggested solution in providing energy-efficient tactile assurances and identifying anomalies in traffic.

Query-based 3D object detection has gained significant success in the application of autonomous driving due to its ability to achieve good performance while maintaining low computational cost. However, it still struggles with the reliable detection of small objects such as bicycles and pedestrians. To address this challenge, this paper introduces a novel sparse query-based approach, termed DinoQuery. This approach utilizes Grounding-DINO with textual prompts to select small-sized objects and generate 2D category-aware queries. These 2D category-aware queries combined with 2D global queries are then lifted to 3D queries by associating each sampled query with its respective 3D position, orientation, and size. The validity of these 3D queries, along with the 2D queries, is verified by the Comprehensive Contrastive Learning (CCL) mechanism. This is achieved by aligning all 2D and 3D queries with their respective 2D and 3D ground truth labels, and computing similarity to select true positive and false positive queries. Then a contrastive loss is introduced to enhance true positive queries and weaken false positive ones based on geometric and semantic similarity. The DinoQuery was tested on the nuScenes dataset and demonstrated excellent performance. Notably, the largest increase of our method is 3.2% on NDS and 3.1% on mAP.

A possible approach to address the ossification of the Internet is to set up multiple virtual networks on top of a single shared physical substrate, and each virtual network can be customized to some specific purposes. However, the current realization of virtualized infrastructure, more specifically, the IaaS cloud and network testbed, suffer from the limited programmability or high cost. To overcome these drawbacks, we propose SDViNet, a solution combining the network virtualization concept applied in IaaS cloud and SDN (Software Defined Network)-based networking testbed. It aims to achieve flexible programmable virtual network abstraction, extend the system at scale and underlying technology easily, and utilize the existing cloud platforms and commercial devices at low-cost. SDViNet is comprised of a centralized server, a set of distributed agents and a central OpenFlow controller, which communicates with the underlying packet delivery fabric. Our prototype implementation and case study show that it is feasible to provide a programmable virtual network in the IaaS cloud, while utilizing the virtualized infrastructure in an extensible and lost-cost way.