4 December 2018 in IEEE Access
1 December 2019 in Neurocomputing
7 October 2021 in International Journal of Crowd Science
Purpose: In recent years, with the increase in computing power, artificial intelligence can gradually be regarded as intelligent agents and interact with humans, this interactive network has become increasingly complex. Therefore, it is necessary to model and analyze this complex interactive network. This paper aims to model and demonstrate the evolution of crowd intelligence using visual complex networks. Design/methodology/approach: This paper uses the complex network to model and observe the collaborative evolution behavior and self-organizing system of crowd intelligence. Findings: The authors use the complex network to construct the cooperative behavior and self-organizing system in crowd intelligence. Determine the evolution mode of the node by constructing the interactive relationship between nodes and observe the global evolution state through the force layout. Practical implications: The simulation results show that the state evolution map can effectively simulate the distribution, interaction and evolution of crowd intelligence through force layout and the intelligent agents’ link mode the authors proposed. Originality/value: Based on the complex network, this paper constructs the interactive behavior and organization system in crowd intelligence and visualizes the evolution process.
22 May 2026 in ACM Transactions on Multimedia Computing, Communications, and Applications
Standard image codecs are misaligned with the feature space of vision-language models like CLIP, because they are designed to exploit spatial redundancies in pixel grids. In contrast, these features are dominated by channel-wise redundancy and contain only limited spatial structure. We empirically find that ViT-B/32 and ViT-B/16 features exhibit weak yet non-negligible spatial redundancy, sufficient to benefit from light local modeling but not from the aggressive spatial downsampling commonly used in learned image codecs. Guided by this, we develop an end-to-end feature compression framework for ViT-based CLIP. A compact transform maintains essential local structure while focusing modeling capacity on channel dependencies and entropy efficiency. Beyond the codec, we present a distortion-aware post-training strategy that adjusts CLIP to compression-related perturbations. A min–max optimization objective enhances robustness against worst-case feature distortions, and a comparison with an MSE-based post-training variant highlights the benefit of the robust formulation. Experiments on Flickr30k retrieval with \(224\times 224\) inputs show that our method consistently outperforms both traditional codecs (e.g., BPG and VVC intra) and learning based codecs (e.g., MLIC++ and STF), even without post-training. Additional post-training further increases this advantage across bitrates. Additional results with ViT-B/16, latent-space visualizations, kernel-size and stride ablations, and CPU inference measurements support the effectiveness and practicality of the proposed framework.