Zijun Long

Quantitative Analysis of the Effects of Irrigation Frequency Under Constant Total Irrigation Amount on Photosynthetic Accumulation, Source–Sink Coordination, and Water–Grain–Quality Synergy in Wide-Row Precision-Sown Winter Wheat

Shengfeng Wang, Enlai Zhan, Guowei Liang, Zijun Long et autres

To address the issues of low water and fertilizer use efficiency and limited yield potential in traditional winter wheat cultivation in Henan Province, and to determine the optimal drip irrigation frequency for wide-row precision sowing under a constant total irrigation amount, this …

cn (code pays fourni par la source)

Regionally Routed Attention U-Net With Latent Fading Kalman Filtering for Real-Time EIT

Tianyi Deng, Zhaosheng Teng, Shin Takahashi, Bojun Liu et autres

Real-time electrical impedance tomography requires reconstruction algorithms that reconcile high structural fidelity with strict latency constraints. We propose RNU-Net, a high-performance, end-to-end time-difference EIT reconstruction framework based on a specialized 1-D ConvNeXt architecture augmented with regionally routed attention. To enable robust mapping, …

cn, jp (code pays fourni par la source)

ADaFuSE: Adaptive Diffusion-generated Image and Text Fusion for Interactive Text-to-Image Retrieval

Zhuocheng Zhang, Xingwu Zhang, Kangheng Liang, Guanxuan Li et autres

Recent advances in interactive text-to-image retrieval (I-TIR) use diffusion models to bridge the modality gap between the textual information need and the images to be searched, resulting in increased effectiveness. However, existing frameworks fuse multi-modal views of user feedback by simple embedding …

ADaFuSE: Adaptive Diffusion-generated Image and Text Fusion for Interactive Text-to-Image Retrieval

Zhuocheng Zhang, Xingwu Zhang, Kangheng Liang, Guanxuan Li et autres

Recent advances in interactive text-to-image retrieval (I-TIR) use diffusion models to bridge the modality gap between the textual information need and the images to be searched, resulting in increased effectiveness. However, existing frameworks fuse multi-modal views of user feedback by simple embedding …

cn, gb (code pays fourni par la source)

Is Task-Specific Training Necessary for Anomaly Detection?

Xingwu Zhang, Guanxuan Li, Paul Henderson, Gerardo Aragon-Camarasa et autres

Current state-of-the-art multi-class unsupervised anomaly detection (MUAD) methods rely on training encoder--decoder models to reconstruct anomaly-free features. However, we argue that such task-specific training is costly under distribution shifts, and that reconstruction-based residual scoring further faces a fidelity--stability dilemma. Existing training-free alternatives, …

Is Task-Specific Training Necessary for Anomaly Detection?

Xingwu Zhang, Guanxuan Li, Paul Henderson, Gerardo Aragon-Camarasa et autres

Current state-of-the-art multi-class unsupervised anomaly detection (MUAD) methods rely on training encoder--decoder models to reconstruct anomaly-free features. However, we argue that such task-specific training is costly under distribution shifts, and that reconstruction-based residual scoring further faces a fidelity--stability dilemma. Existing training-free alternatives, …

Eliminating Hallucination in Diffusion-Augmented Interactive Text-to-Image Retrieval

Zhuocheng Zhang, Kangheng Liang, Guanxuan Li, Paul Henderson et autres

Diffusion-Augmented Interactive Text-to-Image Retrieval (DAI-TIR) is a promising paradigm that improves retrieval performance by generating query images via diffusion models and using them as additional ``views'' of the user's intent. However, these generative views can be incorrect because diffusion generation may introduce …

Eliminating Hallucination in Diffusion-Augmented Interactive Text-to-Image Retrieval

Zhuocheng Zhang, Kangheng Liang, Guanxuan Li, Paul Henderson et autres

Diffusion-Augmented Interactive Text-to-Image Retrieval (DAI-TIR) is a promising paradigm that improves retrieval performance by generating query images via diffusion models and using them as additional ``views'' of the user's intent. However, these generative views can be incorrect because diffusion generation may introduce …

cn, gb (code pays fourni par la source)

Drip Irrigation Coupled with Wide-Row Precision Seeding Enhances Winter Wheat Yield and Water Use Efficiency by Optimizing Canopy Structure and Photosynthetic Performance

Shengfeng Wang, Enlai Zhan, Zijun Long, Guowei Liang et autres

To address the bottlenecks of low water and fertilizer utilization efficiency and limited yield potential inherent in Henan Province’s traditional winter wheat cultivation model of “furrow irrigation + conventional row seeding”, this study delved into the synergistic regulatory mechanisms of drip irrigation …

cn (code pays fourni par la source)

RoboEye: Enhancing 2D Robotic Object Identification with Selective 3D Geometric Keypoint Matching

Xingwu Zhang, Guanxuan Li, Zhuocheng Zhang, Zijun Long

The rapidly growing number of product categories in large-scale e-commerce makes accurate object identification for automated packing in warehouses substantially more difficult. As the catalog grows, intra-class variability and a long tail of rare or visually similar items increase, and when combined …

Optimization Research on Diameter and Speed Matching for a Rim-Driven Thruster

Zichen Song, Zhuo Zhang, Wu Ouyang, Zijun Long et autres

As research hotspot of high-performance electric propulsion devices for ships, the efficiency improvement of rim-driven thrusters (RDTs) is of great significance for their application and promotion. However, the matching of diameter and rotational speed of RDTs is often neglected in the design …

cn, be (code pays fourni par la source)

Diffusion Augmented Retrieval: A Training-Free Approach to Interactive Text-to-Image Retrieval

Zijun Long, Gerardo Aragón-Camarasa, Richard McCreadie, Paul D. Henderson

Interactive Text-to-image retrieval (I-TIR) is an important enabler for a wide range of state-of-the-art services in domains such as e-commerce and education.However, current methods rely on finetuned Multimodal Large Language Models (MLLMs), which are costly to train and update, and exhibit poor …

cn, gb (code pays fourni par la source)