JCJiabao Cui

Jiabao Cui

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Jiabao Cui has a total of 15 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    A Local-Density Structural Developmental Neural Network for Online Incremental Clustering

    by Zhiyong Ding, Jiabao Cui, Xin Xu, Wenzhuo Zhang, Tenglong Liu, Xinwang Liu

    Discovering the evolving class structure of a data stream is a prerequisite for autonomous systems in open environments, where samples arrive one at a time, cannot be revisited, lack annotations, and belong to an unknown, time-varying number of classes. Clustering such streams demands both accurate local distribution modeling and reliable global cluster structure identification within single-pass processing, yet reconciling both objectives remains challenging for existing methods. This paper proposes the local-density structural developmental neural network (LDSDNN), which unifies the two aspects within the developmental process of neurons. Each neuron models its region as an anisotropic Gaussian whose ellipsoidal coverage, defined through the Mahalanobis distance and χ² confidence regions, fits clusters of arbitrary orientation and eccentricity. Inspired by lateral inhibition in biological nervous systems, global structure identification is recast as an online frequency competition in which only neurons at genuine density peaks mature into confirmed clusters, while a volume-based merging mechanism consolidates redundant neurons arising from incremental covariance estimation. Experiments on two synthetic and three real-world datasets show that LDSDNN improves test-set accuracy by up to 26.1 percentage points over the best competing single-phase method, runs one to two orders of magnitude faster than two-phase methods at comparable accuracy, and maintains a compact network whose size tracks the intrinsic complexity of the data.

  • Epoch-Evolving Gaussian Process Guided Learning for Classification

    by Jiabao Cui, Xuewei Li, Hanbin Zhao, Hui Wang, Bin Li, Xi Li

    The conventional mini-batch gradient descent algorithms are usually trapped in the local batch-level distribution information, resulting in the ``zig-zag'' effect in the learning process. To characterize the correlation information between the batch-level distribution and the global data distribution, we propose a novel learning scheme called epoch-evolving Gaussian process guided learning (GPGL) to encode the global data distribution information in a non-parametric way. Upon a set of class-aware anchor samples, our GP model is built to estimate the class distribution for each sample in mini-batch through label propagation from the anchor samples to the batch samples. The class distribution, also named the context label, is provided as a complement for the ground-truth one-hot label. Such a class distribution structure has a smooth property and usually carries a rich body of contextual information that is capable of speeding up the convergence process. With the guidance of the context label and ground-truth label, the GPGL scheme provides a more efficient optimization through updating the model parameters with a triangle consistency loss. Furthermore, our GPGL scheme can be generalized and naturally applied to the current deep models, outperforming the state-of-the-art optimization methods on six benchmark datasets.


  • Multitask Identity-Aware Image Steganography via Minimax Optimization

    by Jiabao Cui, Pengyi Zhang, Songyuan Li, Liangli Zheng, Cuizhu Bao, Jupeng Xia, Xi Li

    High-capacity image steganography, aimed at concealing a secret image in a cover image, is a technique to preserve sensitive data, e.g., faces and fingerprints. Previous methods focus on the security during transmission and subsequently run a risk of privacy leakage after the restoration of secret images at the receiving end. To address this issue, we propose a framework, called Multitask Identity-Aware Image Steganography (MIAIS), to achieve direct recognition on container images without restoring secret images. The key issue of the direct recognition is to preserve identity information of secret images into container images and make container images look similar to cover images at the same time. Thus, we introduce a simple content loss to preserve the identity information, and design a minimax optimization to deal with the contradictory aspects. We demonstrate that the robustness results can be transferred across different cover datasets. In order to be flexible for the secret image restoration in some cases, we incorporate an optional restoration network into our method, providing a multitask framework. The experiments under the multitask scenario show the effectiveness of our framework compared with other visual information hiding methods and state-of-the-art high-capacity image steganography methods. The code is available at https://github.com/jiabaocui/MIAIS .