EZEnyuan Zhao

Enyuan Zhao

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Enyuan Zhao has a total of 25 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • RMPT: Retrieval-based multimodal prompt tuning for event detection

    by Enyuan Zhao, Jie Nie, Lei Huang, Ning Song, Zhiqiang Wei


  • Open Access

    Frequency Spectrum Adaptor for Remote Sensing Image–Text Retrieval

    by Ziyi Wan, Enyuan Zhao, Jie Nie, Ze Zhang, Zhiqiang Wei, Nan Zheng, Yuting Zhao

    Remote sensing image–text retrieval (RSITR) is a critical task that involves parsing the content of remote sensing (RS) images to match semantically relevant text. Existing RSITR methods primarily focus on directly adopting pretrained models and performing transfer learning through fine-tuning, neglecting the complex high-dimensional structured information present in RS images, where texture, color, scale, and semantics are tightly coupled. Consequently, these methods exhibit limitations in handling specific receptive fields or preserving the structural information within RS images, leading to inaccurate retrieval matches. To mitigate this issue, this article proposes a frequency spectrum adapter for RSITR, which aims to perceive the unique structured information of RS images to facilitate the transfer of visual–linguistic knowledge from the natural domain to the RS domain. The main contributions of this article are as follows. First, a frequency-domain-based RS image–text retrieval adapter (FRS-Adapter) was developed. By expanding the spectral receptive field, it extracts the unique structured information of RS images, enhancing the fine-tuning effect of RS to natural scene domain transfer. Second, a unimodal filter bank was designed, which uses filter banks for unimodal spectral compression across multiple bands. Within each band, the spectral features of amplitude and phase structures are utilized to further enhance the representation of structured information. Third, a cross-modal spectrum mutual aggregation module was introduced to promote the deep integration of linguistic, spatial, and spectral information. This guides the retention of relevant frequency components and effectively reduces the impact of irrelevant frequency components. Fourth, we conduct quantitative and qualitative experiments on three large RS cross-modal retrieval datasets, validating the significant performance of the FRS-Adapter in RSITR.


  • Frequency domain transfer learning for remote sensing visual question answering

    by Ze Zhang, Enyuan Zhao, Ziyi Wan, Xinyue Liang, Min Ye, Jie Nie, Lei Huang


  • Open Access

    Language-guided Bias Generation Contrastive Strategy for Visual Question Answering

    by Enyuan Zhao, Ning Song, Ze Zhang, Jie Nie, Xinyue Liang, Zhiqiang Wei

    Visual question answering (VQA) is a challenging task that requires models to understand both visual and linguistic inputs and produce accurate answers. However, VQA models often exploit biases in datasets to make predictions rather than reasoning based on the inputs. Prior approaches to debiasing have suggested the implementation of a supplementary model, deliberately designed to exhibit bias, which subsequently informs the training of a resilient target model. Nevertheless, such techniques merely quantify the model’s divergence based on the statistical distribution of labels within the training dataset or in relation to unimodal branches. In this work, we propose a novel method of generating bias from the target model itself, called LEGO, which aims to combat the language guidance bias. Specifically, LEGO framework employs a generative network that assimilates the biases inherent in the target model by integrating adversarial goals with the principles of knowledge distillation. Then, we use a debiased contrastive learning strategy to model the language guidance bias of caption and question. In the process of modeling, in order to obtain robust semantic coreference, the multimodal representations of two semantic granularity are modeled by mutual information fusion and contrast learning difference modeling. We evaluate our method on various VQA-biased datasets, including VQA-CP2, GQA-OOD, and RSICD, and show that it outperforms similar methods.


  • A Spatial-Frequency Fusion Strategy Based on Linguistic Query Refinement for RSVG

    by Enyuan Zhao, Ziyi Wan, Ze Zhang, Jie Nie, Xinyue Liang, Lei Huang

    Remote Sensing Visual Grounding (RSVG) represents a pivotal task aimed at pinpointing regions highly relevant to textual descriptions by parsing remote sensing image content. Existing RSVG methodologies focus on multimodal semantic integration on specific datasets, overlooking the intricate high-dimensional information present in remote sensing images, where color, scale, and semantics are tightly coupled. Consequently, these approaches exhibit limitations in handling specific receptive fields or preserving structural information within remote sensing images, leading to insufficient localization precision. To address this issue, this paper introduces a strategy that incorporates spatial frequency information, leveraging Fourier transforms to capture global structured information from remote sensing data. This is followed by a progressive aggregation across spatial, spectral, and linguistic modalities to achieve robust semantic coreference. The primary contributions of this paper are as follows: Firstly, a spatial-frequency fusion strategy based on linguistic query refinement is proposed, which enhances visual grounding performance significantly by extracting spectral features with potent spatial perception capabilities through the expansion of the spectral receptive field. Secondly, a frequency-guided spatial module is designed, utilizing amplitude and phase-structured spectral features to augment spatial representation capabilities further. Lastly, a query-aware original attention mechanism is developed, facilitating deep integration of spatial and spectral information under linguistic guidance. Extensive experimentation on the RSVGD dataset validates the efficacy of the proposed approach, demonstrating superior performance when compared to state-of-the-art methods.


  • Open Access

    Intelligent fitting global real‐time task scheduling strategy for high‐performance multi‐core systems

    by Junpeng Wu, Enyuan Zhao, Sizhao Li, Yanqiang Wang

    With the development of high‐performance computing, it is possible to solve large‐scale computing problems. However, the irregularity and access characteristics of computing problems bring challenges to the realisation and performance optimisation. Improving the performance of a single core makes it challenging to maintain Moore's law, and multi‐core processors emerge. A chip brings together multiple universal processor cores of equal status and has the same structure supported by an isomorphic multi‐core processor. In high‐performance computing, the granularity of computing tasks leads to the complexity of scheduling strategies. Satisfying high system performance, load balancing and processor fault tolerance at a minimum cost is the key to task scheduling in the high‐performance field, especially in specific multi‐core hardware architecture. In this study, global real‐time task scheduling is implemented in a high‐performance multi‐core system. The system adopts the hybrid scheduling among clusters and the intelligent fitting within clusters to implement the global real‐time task scheduling strategy. In the cluster scheduling policy, tasks are allowed to preempt the core with low priority, and the priority of tasks that access memory is dynamically improved, higher than that of all the tasks without memory access. An intelligent fitting method is also proposed. When the data read by the task is in the cache and the cache access ability value of the task is within a reasonable threshold, the priority of the task is promoted to the highest priority, preempting the core without the access memory task. The results show that the intelligently fitting global scheduling strategy for multi‐core systems has better performance in the nuclear utilisation rate and task schedulability.