Fan Zhao
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
Research Subjects
Fan Zhao has a total of 69 co-authors. Here are the top co-authors by number of shared publications:
- Guangfeng Lin
Information Science Department, Xi'an University of Technology, Xi'an, 710048, Shaanxi Province, China.
5 shared publications - Yajun Chen
Department of Information Science, Xi’an University of Technology, Xi’an 710048, China
3 shared publications Erhu Zhang
Xi'an Univ. of Technology (China)
2 shared publicationsFangying Qu
Georgia Institute of Technology, Atlanta, USA
2 shared publicationsKaidi Hui
Department of Information Science, Xi’an University of Technology, Xi’an, China
2 shared publications- Kaiyang Liao
School of Printing, Packaging and Digital Media, Xi’an University of Technology, Xi’an, China
2 shared publications
Latest Publications
- 2 September 2026
- Published by Springer Nature
- p. 209-220
Leakage Period Localization Method for Water Distribution Network Based on Continuous-Time Dynamic Graph
- 2 September 2026
- Published by Springer Nature
- p. 209-220
Leakage Period Localization Method for Water Distribution Network Based on Continuous-Time Dynamic Graph
- Open Access
- 9 February 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-13
Cross modal recipe retrieval with fine grained modal interaction
Designing systems capable of finding relevant cooking recipes given a user-submitted food image, or vice versa, cross-modal recipe retrieval has gained significant attention in recent years. Numerous advanced techniques have been employed to improve the performance of cross-modal recipe retrieval on general benchmarks. However, leveraging the fine-grained modalities interaction for enhancing multi-modal representation is still limited. Preceding a hierarchical recipe Transformer for encoding individual recipe components, we introduce the cross-component multiscale recipe enriching (CCMRE) module, which enhances the components of the recipe through fully convolutional operations with convolutional kernels of different lengths. Further, we embed a text-contextualized visual enhancing (TCVE) module into an intermediate layer of the image encoder to enrich the visual encoder. Utilizing the similarity between image local features and intermediate recipe representations, TCVE enhances visual representation by deeper model relearning. We conduct a thorough analysis and ablation studies to validate the proposed method, FMI (Fine-grained Modalities Interaction for Cross-Modal Recipe Retrieval). As a result, our method outperforms current SoTA across all metrics on the Recipe1M dataset. Specifically, compared to baseline model, the improvements of + 17.4 R@1 and + 20.5R@1 on the 1 k and 10 k test sets are achieved respectively.
- 9 February 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-13
Open AccessCross modal recipe retrieval with fine grained modal interaction
Designing systems capable of finding relevant cooking recipes given a user-submitted food image, or vice versa, cross-modal recipe retrieval has gained significant attention in recent years. Numerous advanced techniques have been employed to improve the performance of cross-modal recipe retrieval on general benchmarks. However, leveraging the fine-grained modalities interaction for enhancing multi-modal representation is still limited. Preceding a hierarchical recipe Transformer for encoding individual recipe components, we introduce the cross-component multiscale recipe enriching (CCMRE) module, which enhances the components of the recipe through fully convolutional operations with convolutional kernels of different lengths. Further, we embed a text-contextualized visual enhancing (TCVE) module into an intermediate layer of the image encoder to enrich the visual encoder. Utilizing the similarity between image local features and intermediate recipe representations, TCVE enhances visual representation by deeper model relearning. We conduct a thorough analysis and ablation studies to validate the proposed method, FMI (Fine-grained Modalities Interaction for Cross-Modal Recipe Retrieval). As a result, our method outperforms current SoTA across all metrics on the Recipe1M dataset. Specifically, compared to baseline model, the improvements of + 17.4 R@1 and + 20.5R@1 on the 1 k and 10 k test sets are achieved respectively.
- 6 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (11) , 17987-18003
High-Order Structure-Preserving Graph Neural Network for Few-Shot Learning
Few-shot learning can find the latent structure information between the support and query data by the similarity metric of meta-learning to construct the discriminative model for recognizing the new categories with the less labeled samples. Most existing methods attempt to model the similarity relationships among samples within meta-tasks for achieving this goal. However, the relationship assessment among samples from distinct meta-tasks is challenging due to the differing metric relationships inherent to each respective meta-task. To address this issue, high-order structure preserving graph neural network (HOSP-GNN) can deeply explore the rich samples structure of the different meta-tasks to predict the label of the queried data based on the graph. HOSP-GNN can mine high-order structures to enhance their relevance by triple samples. In addition, it can also generate the updating rule of the different-order structures for node and edge representation optimization under manifold constraints. Notably, HOSP-GNN eliminates the need for retraining the learning model to recognize new classes, thanks to its high generalization of high-order structure that ensures model adaptability. The experiments demonstrate that HOSP-GNN outperforms state-of-the-art methods in four benchmark datasets, as well as on a self-built dataset focusing on endangered animals. The available code is https://github.com/yangfeifei02/HOSP.
- 6 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (11) , 17987-18003
High-Order Structure-Preserving Graph Neural Network for Few-Shot Learning
Few-shot learning can find the latent structure information between the support and query data by the similarity metric of meta-learning to construct the discriminative model for recognizing the new categories with the less labeled samples. Most existing methods attempt to model the similarity relationships among samples within meta-tasks for achieving this goal. However, the relationship assessment among samples from distinct meta-tasks is challenging due to the differing metric relationships inherent to each respective meta-task. To address this issue, high-order structure preserving graph neural network (HOSP-GNN) can deeply explore the rich samples structure of the different meta-tasks to predict the label of the queried data based on the graph. HOSP-GNN can mine high-order structures to enhance their relevance by triple samples. In addition, it can also generate the updating rule of the different-order structures for node and edge representation optimization under manifold constraints. Notably, HOSP-GNN eliminates the need for retraining the learning model to recognize new classes, thanks to its high generalization of high-order structure that ensures model adaptability. The experiments demonstrate that HOSP-GNN outperforms state-of-the-art methods in four benchmark datasets, as well as on a self-built dataset focusing on endangered animals. The available code is https://github.com/yangfeifei02/HOSP.
- 12 November 2024
- Published by Springer Nature
- p. 79-90
Fine-Grained Modalities Interaction for Cross-Modal Recipe Retrieval
- 12 November 2024
- Published by Springer Nature
- p. 79-90
Fine-Grained Modalities Interaction for Cross-Modal Recipe Retrieval
- Open Access
- 1 April 2024
- Published by American Dairy Science Association in Journal of Dairy Science
- Vol. 107 (4) , 1928-1949
Enhancing the antioxidant capacity and quality attributes of fermented goat milk through the synergistic action of Limosilactobacillus fermentum WXZ 2-1 with a starter culture
This study evaluated 75 strains of lactic acid bacteria (LAB) isolated from traditional dairy products in western China for their probiotic properties. Among them, Limosilactobacillus fermentum WXZ 2-1, Lactiplantibacillus plantarum TXZ 2-35, Companilactobacillus crustorum QHS 9, and Companilactobacillus crustorum QHS 10 demonstrated potential probiotic characteristics. The antioxidant capacity of these 4 strains was assessed, revealing that L. fermentum WXZ 2-1 exhibited the highest antioxidant capacity. Furthermore, when cocultured with Streptococcus salivarius ssp. thermophiluss and Lactobacillus delbrueckii ssp. bulgaricus, L. fermentum WXZ 2-1 demonstrated a synergistic effect in growth medium and goat milk. To explore its impact on goat milk fermentation, different amounts of L. fermentum WXZ 2-1 were added to goat milk, and its physicochemical properties, antioxidant activity, flavor substances, and metabolomics were analyzed. The study found that the incorporation of L. fermentum WXZ 2-1 in goat milk fermentation significantly improved the texture characteristics, antioxidant capacity, and flavor of fermented goat milk. These findings highlight the potential of L. fermentum WXZ 2-1 as a valuable probiotic strain for enhancing the functionality and desirability of fermented goat milk, contributing to the development of novel functional foods with improved health benefits and enhanced quality attributes.
- 1 April 2024
- Published by American Dairy Science Association in Journal of Dairy Science
- Vol. 107 (4) , 1928-1949
Open AccessEnhancing the antioxidant capacity and quality attributes of fermented goat milk through the synergistic action of Limosilactobacillus fermentum WXZ 2-1 with a starter culture
This study evaluated 75 strains of lactic acid bacteria (LAB) isolated from traditional dairy products in western China for their probiotic properties. Among them, Limosilactobacillus fermentum WXZ 2-1, Lactiplantibacillus plantarum TXZ 2-35, Companilactobacillus crustorum QHS 9, and Companilactobacillus crustorum QHS 10 demonstrated potential probiotic characteristics. The antioxidant capacity of these 4 strains was assessed, revealing that L. fermentum WXZ 2-1 exhibited the highest antioxidant capacity. Furthermore, when cocultured with Streptococcus salivarius ssp. thermophiluss and Lactobacillus delbrueckii ssp. bulgaricus, L. fermentum WXZ 2-1 demonstrated a synergistic effect in growth medium and goat milk. To explore its impact on goat milk fermentation, different amounts of L. fermentum WXZ 2-1 were added to goat milk, and its physicochemical properties, antioxidant activity, flavor substances, and metabolomics were analyzed. The study found that the incorporation of L. fermentum WXZ 2-1 in goat milk fermentation significantly improved the texture characteristics, antioxidant capacity, and flavor of fermented goat milk. These findings highlight the potential of L. fermentum WXZ 2-1 as a valuable probiotic strain for enhancing the functionality and desirability of fermented goat milk, contributing to the development of novel functional foods with improved health benefits and enhanced quality attributes.
- 29 December 2023
- Published by Springer Nature in The Visual Computer
- Vol. 40 (9) , 6639-6654
A pseudo-color image-based cylindrical object surface text detection method
- 29 December 2023
- Published by Springer Nature in The Visual Computer
- Vol. 40 (9) , 6639-6654
A pseudo-color image-based cylindrical object surface text detection method
- Open Access
A Crack Segmentation Model Combining Morphological Network and Multiple Loss Mechanism
With the wide application of computer vision technology and deep-learning theory in engineering, the image-based detection of cracks in structures such as pipelines, pavements and dams has received more and more attention. Aiming at the problems of high cost, low efficiency and poor detection accuracy in traditional crack detection methods, this paper proposes a crack segmentation network by combining a morphological network and a multiple-loss mechanism. First, for improving the identification of cracks with different resolutions, the U-Net network is used to extract multi-scale features from the crack image. Second, for eliminating the effect of polarized light on the cracks under different illuminations, the extracted crack features are further morphologically processed by a white-top hat transform and a black-bottom hat transform. Finally, a multi-loss mechanism is designed to solve the problem of the inaccurate segmentation of cracks on a single scale. Extensive experiments are carried out on five open crack datasets: Crack500, CrackTree200, CFD, AEL and GAPs384. The experimental results showed that the average ODS, OIS, AIU, sODS and sOIS are 75.7%, 73.9%, 36.4%, 52.4% and 52.2%, respectively. Compared with state-of-the-art methods, the proposed method achieves better crack segmentation performance. Ablation experiments also verified the effectiveness of each module in the algorithm.
Open AccessA Crack Segmentation Model Combining Morphological Network and Multiple Loss Mechanism
With the wide application of computer vision technology and deep-learning theory in engineering, the image-based detection of cracks in structures such as pipelines, pavements and dams has received more and more attention. Aiming at the problems of high cost, low efficiency and poor detection accuracy in traditional crack detection methods, this paper proposes a crack segmentation network by combining a morphological network and a multiple-loss mechanism. First, for improving the identification of cracks with different resolutions, the U-Net network is used to extract multi-scale features from the crack image. Second, for eliminating the effect of polarized light on the cracks under different illuminations, the extracted crack features are further morphologically processed by a white-top hat transform and a black-bottom hat transform. Finally, a multi-loss mechanism is designed to solve the problem of the inaccurate segmentation of cracks on a single scale. Extensive experiments are carried out on five open crack datasets: Crack500, CrackTree200, CFD, AEL and GAPs384. The experimental results showed that the average ODS, OIS, AIU, sODS and sOIS are 75.7%, 73.9%, 36.4%, 52.4% and 52.2%, respectively. Compared with state-of-the-art methods, the proposed method achieves better crack segmentation performance. Ablation experiments also verified the effectiveness of each module in the algorithm.
- Open Access
- 10 June 2022
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 82 (1) , 879-898
A coarse-to-fine temporal action detection method combining light and heavy networks
Temporal action detection aims to judge whether there existing a certain number of action instances in a long untrimmed videos and to locate the start and end time of each action. Even though the existing action detection methods have shown promising results in recent years with the widespread application of Convolutional Neural Network (CNN), it is still a challenging problem to accurately locate each action segment while ensuring real-time performance. In order to achieve a good tradeoff between detection efficiency and accuracy, we present a coarse-to-fine hierarchical temporal action detection method by using multi-scale sliding window mechanism. Since the complexity of the convolution operator is proportional to the number and the size of the input video clips, the idea of our proposed method is to first determine candidate action proposals and then perform the detection task on these candidate action proposals only with a view to reducing the overall complexity of the detection method. By making full use of the spatio-temporal information of video clips, a lightweight 3D-CNN classifier is first used to quickly determine whether the video clip is a candidate action proposal, avoiding the re-detection of a large number of non-action video clips by the heavyweight deep network. A heavyweight detector is designed to further improve the accuracy of action positioning by considering both boundary regression loss and category loss in the target loss function. In addition, the Non-Maximum Suppression (NMS) is performed to eliminate redundant detection results among the overlapping proposals. The mean Average Precision (mAP) is 40.6%, 51.7% and 20.4% on THUMOS14, ActivityNet and MPII Cooking dataset when the Intersection-over-Union (tIoU) threshold is set to 0.5, respectively. Experimental results show the superior performance of the proposed method on three challenging temporal activity detection datasets while achieving real-time speed. At the same time, our method can generate proposals for unseen action classes with high recalls.
- 10 June 2022
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 82 (1) , 879-898
Open AccessA coarse-to-fine temporal action detection method combining light and heavy networks
Temporal action detection aims to judge whether there existing a certain number of action instances in a long untrimmed videos and to locate the start and end time of each action. Even though the existing action detection methods have shown promising results in recent years with the widespread application of Convolutional Neural Network (CNN), it is still a challenging problem to accurately locate each action segment while ensuring real-time performance. In order to achieve a good tradeoff between detection efficiency and accuracy, we present a coarse-to-fine hierarchical temporal action detection method by using multi-scale sliding window mechanism. Since the complexity of the convolution operator is proportional to the number and the size of the input video clips, the idea of our proposed method is to first determine candidate action proposals and then perform the detection task on these candidate action proposals only with a view to reducing the overall complexity of the detection method. By making full use of the spatio-temporal information of video clips, a lightweight 3D-CNN classifier is first used to quickly determine whether the video clip is a candidate action proposal, avoiding the re-detection of a large number of non-action video clips by the heavyweight deep network. A heavyweight detector is designed to further improve the accuracy of action positioning by considering both boundary regression loss and category loss in the target loss function. In addition, the Non-Maximum Suppression (NMS) is performed to eliminate redundant detection results among the overlapping proposals. The mean Average Precision (mAP) is 40.6%, 51.7% and 20.4% on THUMOS14, ActivityNet and MPII Cooking dataset when the Intersection-over-Union (tIoU) threshold is set to 0.5, respectively. Experimental results show the superior performance of the proposed method on three challenging temporal activity detection datasets while achieving real-time speed. At the same time, our method can generate proposals for unseen action classes with high recalls.
- Open Access
- 2 June 2022
- Published by MDPI AG in Applied Sciences
- Vol. 12 (11) , 5679
Infrared Bird Target Detection Based on Temporal Variation Filtering and a Gaussian Heat-Map Perception Network
Flying bird detection has recently attracted increasing attention in computer vision. However, compared to conventional object detection tasks, it is much more challenging to trap flying birds in infrared videos due to small target size, complex backgrounds, and dim shapes. In order to solve the problem of poor detection performance caused by insufficient feature information of small and dim birds, this paper suggests a method of detecting birds in outdoor environments using image pre-processing and deep learning, called temporal Variation filtering (TVF) and Gaussian heatmap perception network (GHPNet), respectively. TVF separates the dynamic background from moving creatures. Using bird appearance features that are brightest at the center and gradually darker outwards, the size-adaptive Gaussian kernel is used to generate the ground truth of the region of interest (ROI). In order to fuse the features from different scales and to highlight the saliency of the target, the GHPNet network integrates VGG-16 and maximum-no-pooling filterer into a U-Net network. The comparative experiments demonstrate that the proposed method outperforms those that are state-of-the-art in detecting bird targets in real-world infrared images.
- 2 June 2022
- Published by MDPI AG in Applied Sciences
- Vol. 12 (11) , 5679
Open AccessInfrared Bird Target Detection Based on Temporal Variation Filtering and a Gaussian Heat-Map Perception Network
Flying bird detection has recently attracted increasing attention in computer vision. However, compared to conventional object detection tasks, it is much more challenging to trap flying birds in infrared videos due to small target size, complex backgrounds, and dim shapes. In order to solve the problem of poor detection performance caused by insufficient feature information of small and dim birds, this paper suggests a method of detecting birds in outdoor environments using image pre-processing and deep learning, called temporal Variation filtering (TVF) and Gaussian heatmap perception network (GHPNet), respectively. TVF separates the dynamic background from moving creatures. Using bird appearance features that are brightest at the center and gradually darker outwards, the size-adaptive Gaussian kernel is used to generate the ground truth of the region of interest (ROI). In order to fuse the features from different scales and to highlight the saliency of the target, the GHPNet network integrates VGG-16 and maximum-no-pooling filterer into a U-Net network. The comparative experiments demonstrate that the proposed method outperforms those that are state-of-the-art in detecting bird targets in real-world infrared images.