Hong Wu
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
Hong Wu has a total of 53 co-authors. Here are the top co-authors by number of shared publications:
- Bodi Yuan
School of Automation, Xian University of Posts and Telecommunications, Xian, China
2 shared publications - David Clifford
CSIRO Mathematical and Information Sciences, Locked Bag 17, North Ryde, NSW 1670, Australia
2 shared publications Fiseha B. Tesema
2 shared publications- Hao Su
Osaka University
2 shared publications Junpeng Lin
2 shared publicationsMingjian Chen
2 shared publications
Latest Publications
- 27 October 2026
- Published by Springer Nature in Applied Intelligence
- Vol. 56 (16) , 520
IDB-FOIL: iterative dual-branch inductive logic learning for automated and interpretable image labeling
- 27 October 2026
- Published by Springer Nature in Applied Intelligence
- Vol. 56 (16) , 520
IDB-FOIL: iterative dual-branch inductive logic learning for automated and interpretable image labeling
- 6 May 2026
- Published by Springer Nature in Multimedia Systems
- Vol. 32 (4) , 290
Tabamba: a Mamba-Transformer hybrid model for table structure recognition
- 6 May 2026
- Published by Springer Nature in Multimedia Systems
- Vol. 32 (4) , 290
Tabamba: a Mamba-Transformer hybrid model for table structure recognition
- 17 June 2024
- Published by Springer Nature in Multimedia Systems
- Vol. 30 (4) , 1-9
SenseMLP: a parallel MLP architecture for sensor-based human activity recognition
- 17 June 2024
- Published by Springer Nature in Multimedia Systems
- Vol. 30 (4) , 1-9
SenseMLP: a parallel MLP architecture for sensor-based human activity recognition
- Open Access
- 8 May 2024
- Published by Springer Nature in Neural Processing Letters
- Vol. 56 (3) , 1-17
A Prototype-Based Neural Network for Image Anomaly Detection and Localization
Image anomaly detection and localization perform not only image-level anomaly classification but also locate pixel-level anomaly regions. Recently, it has received much research attention due to its wide application in various fields. This paper proposes ProtoAD, a prototype-based neural network for image anomaly detection and localization. First, the patch features of normal images are extracted by a deep network pre-trained on nature images. Then, the prototypes of the normal patch features are learned by non-parametric clustering. Finally, we construct an image anomaly localization network (ProtoAD) by appending the feature extraction network with L2 feature normalization, a $$1\times 1$$ convolutional layer, a channel max-pooling, and a subtraction operation. We use the prototypes as the kernels of the $$1\times 1$$ convolutional layer; therefore, our neural network does not need a training phase and can conduct anomaly detection and localization in an end-to-end manner. Extensive experiments on two challenging industrial anomaly detection datasets, MVTec AD and BTAD, demonstrate that ProtoAD achieves competitive performance compared to the state-of-the-art methods with a higher inference speed. The code and pre-trained models are publicly available at https://github.com/98chao/ProtoAD.- 8 May 2024
- Published by Springer Nature in Neural Processing Letters
- Vol. 56 (3) , 1-17
Open AccessA Prototype-Based Neural Network for Image Anomaly Detection and Localization
Image anomaly detection and localization perform not only image-level anomaly classification but also locate pixel-level anomaly regions. Recently, it has received much research attention due to its wide application in various fields. This paper proposes ProtoAD, a prototype-based neural network for image anomaly detection and localization. First, the patch features of normal images are extracted by a deep network pre-trained on nature images. Then, the prototypes of the normal patch features are learned by non-parametric clustering. Finally, we construct an image anomaly localization network (ProtoAD) by appending the feature extraction network with L2 feature normalization, a $$1\times 1$$ convolutional layer, a channel max-pooling, and a subtraction operation. We use the prototypes as the kernels of the $$1\times 1$$ convolutional layer; therefore, our neural network does not need a training phase and can conduct anomaly detection and localization in an end-to-end manner. Extensive experiments on two challenging industrial anomaly detection datasets, MVTec AD and BTAD, demonstrate that ProtoAD achieves competitive performance compared to the state-of-the-art methods with a higher inference speed. The code and pre-trained models are publicly available at https://github.com/98chao/ProtoAD. - 25 March 2024
- Published by Elsevier in International Journal of Medical Informatics
- Vol. 186, 105425
Predicting intubation for intensive care units patients: A deep learning approach to improve patient management
- 25 March 2024
- Published by Elsevier in International Journal of Medical Informatics
- Vol. 186, 105425
Predicting intubation for intensive care units patients: A deep learning approach to improve patient management
- 18 April 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Systems, Man, and Cybernetics Magazine
- Vol. 9 (2) , 25-38
Addressee Detection Using Facial and Audio Features in Mixed Human–Human and Human–Robot Settings: A Deep Learning Framework
Addressee detection (AD) enables robots to interact smoothly with a human by distinguishing whether it is being addressed. However, this has not been widely explored. The few studies that have explored this area focused on a human-to-human or human-to-robot conversation confined inside a meeting room using gaze and utterance. These works used statistical and rule-based approaches, which tend to depend on specific settings. Further, they did not fully leverage the available audio and visual information or the short-term and long-term segments, and they have not explored combining important conversation cues—the facial and audio features. In addition, no audiovisual spatiotemporal annotated dataset captured in mixed human-to-human and human-to-robot settings is available to support exploring the area using new approaches.
- 18 April 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Systems, Man, and Cybernetics Magazine
- Vol. 9 (2) , 25-38
Addressee Detection Using Facial and Audio Features in Mixed Human–Human and Human–Robot Settings: A Deep Learning Framework
Addressee detection (AD) enables robots to interact smoothly with a human by distinguishing whether it is being addressed. However, this has not been widely explored. The few studies that have explored this area focused on a human-to-human or human-to-robot conversation confined inside a meeting room using gaze and utterance. These works used statistical and rule-based approaches, which tend to depend on specific settings. Further, they did not fully leverage the available audio and visual information or the short-term and long-term segments, and they have not explored combining important conversation cues—the facial and audio features. In addition, no audiovisual spatiotemporal annotated dataset captured in mixed human-to-human and human-to-robot settings is available to support exploring the area using new approaches.
- 18 November 2022
- Published by Springer Nature in Journal of Digital Imaging
- Vol. 36 (2) , 547-561
Feature Aggregation and Refinement Network for 2D Anatomical Landmark Detection
Localization of anatomical landmarks is essential for clinical diagnosis, treatment planning, and research. This paper proposes a novel deep network named feature aggregation and refinement network (FARNet) for automatically detecting anatomical landmarks. FARNet employs an encoder-decoder structure architecture. To alleviate the problem of limited training data in the medical domain, we adopt a backbone network pre-trained on natural images as the encoder. The decoder includes a multi-scale feature aggregation module for multi-scale feature fusion and a feature refinement module for high-resolution heatmap regression. Coarse-to-fine supervisions are applied to the two modules to facilitate end-to-end training. We further propose a novel loss function named Exponential Weighted Center loss for accurate heatmap regression, which focuses on the losses from the pixels near landmarks and suppresses the ones from far away. We evaluate FARNet on three publicly available anatomical landmark detection datasets, including cephalometric, hand, and spine radiographs. Our network achieves state-of-the-art performances on all three datasets. Code is available at https://github.com/JuvenileInWind/FARNet.
- 18 November 2022
- Published by Springer Nature in Journal of Digital Imaging
- Vol. 36 (2) , 547-561
Feature Aggregation and Refinement Network for 2D Anatomical Landmark Detection
Localization of anatomical landmarks is essential for clinical diagnosis, treatment planning, and research. This paper proposes a novel deep network named feature aggregation and refinement network (FARNet) for automatically detecting anatomical landmarks. FARNet employs an encoder-decoder structure architecture. To alleviate the problem of limited training data in the medical domain, we adopt a backbone network pre-trained on natural images as the encoder. The decoder includes a multi-scale feature aggregation module for multi-scale feature fusion and a feature refinement module for high-resolution heatmap regression. Coarse-to-fine supervisions are applied to the two modules to facilitate end-to-end training. We further propose a novel loss function named Exponential Weighted Center loss for accurate heatmap regression, which focuses on the losses from the pixels near landmarks and suppresses the ones from far away. We evaluate FARNet on three publicly available anatomical landmark detection datasets, including cephalometric, hand, and spine radiographs. Our network achieves state-of-the-art performances on all three datasets. Code is available at https://github.com/JuvenileInWind/FARNet.
- Open Access
- 25 October 2022
- Published by Springer Nature in Research Square
A Prototype-Based Neural Network for Image Anomaly Detection and Localization
Image anomaly detection and localization perform not only imagelevel anomaly classification but also locate pixel-level anomaly regions. Recently, it has received much research attention due to its wide application in various fields. This paper proposes ProtoAD, a prototypebased neural network for image anomaly detection and localization. First, the patch features of normal images are extracted by a deep network pre-trained on nature images. Then, the prototypes of the normal patch features are learned by non-parametric clustering. Finally, we construct an image anomaly localization network (ProtoAD) by appending the feature extraction network with the L2 feature normalization, a 1 × 1 convolutional layer, a channel max-pooling, and a subtraction operation. We use the prototypes as the kernels of the 1 × 1 convolutional layer; therefore, our neural network does not need a training phase and can conduct anomaly detection and localization in an end-to-end manner. Extensive experiments on the MVTec AD dataset demonstrate that ProtoAD achieves competitive performance compared to the state-of-the-art methods with a higher inference speed. The source code will be released on GitHub.
- 25 October 2022
- Published by Springer Nature in Research Square
Open AccessA Prototype-Based Neural Network for Image Anomaly Detection and Localization
Image anomaly detection and localization perform not only imagelevel anomaly classification but also locate pixel-level anomaly regions. Recently, it has received much research attention due to its wide application in various fields. This paper proposes ProtoAD, a prototypebased neural network for image anomaly detection and localization. First, the patch features of normal images are extracted by a deep network pre-trained on nature images. Then, the prototypes of the normal patch features are learned by non-parametric clustering. Finally, we construct an image anomaly localization network (ProtoAD) by appending the feature extraction network with the L2 feature normalization, a 1 × 1 convolutional layer, a channel max-pooling, and a subtraction operation. We use the prototypes as the kernels of the 1 × 1 convolutional layer; therefore, our neural network does not need a training phase and can conduct anomaly detection and localization in an end-to-end manner. Extensive experiments on the MVTec AD dataset demonstrate that ProtoAD achieves competitive performance compared to the state-of-the-art methods with a higher inference speed. The source code will be released on GitHub.
- 5 September 2022
- Published by Springer Nature in Multimedia Systems
- Vol. 29 (1) , 261-273
Learning unified anchor graph based on affinity relationships with strong consensus for multi-view spectral clustering
- 5 September 2022
- Published by Springer Nature in Multimedia Systems
- Vol. 29 (1) , 261-273
Learning unified anchor graph based on affinity relationships with strong consensus for multi-view spectral clustering
- Open Access
Semantically Robust Unpaired Image Translation for Data with Unmatched Semantics Statistics
Many applications of unpaired image-to-image translation require the input contents to be preserved semantically during translations. Unaware of the inherently unmatched semantics distributions between source and target domains, existing distribution matching methods (i.e., GAN-based) can give undesired solutions. In particular, although producing visually reasonable outputs, the learned models usually flip the semantics of the inputs. To tackle this without using extra supervision, we propose to enforce the translated outputs to be semantically invariant w.r.t. small perceptual variations of the inputs, a property we call "semantic robustness". By optimizing a robustness loss w.r.t. multi-scale feature space perturbations of the inputs, our method effectively reduces semantics flipping and produces translations that outperform existing methods both quantitatively and qualitatively.
Open AccessSemantically Robust Unpaired Image Translation for Data with Unmatched Semantics Statistics
Many applications of unpaired image-to-image translation require the input contents to be preserved semantically during translations. Unaware of the inherently unmatched semantics distributions between source and target domains, existing distribution matching methods (i.e., GAN-based) can give undesired solutions. In particular, although producing visually reasonable outputs, the learned models usually flip the semantics of the inputs. To tackle this without using extra supervision, we propose to enforce the translated outputs to be semantically invariant w.r.t. small perceptual variations of the inputs, a property we call "semantic robustness". By optimizing a robustness loss w.r.t. multi-scale feature space perturbations of the inputs, our method effectively reduces semantics flipping and produces translations that outperform existing methods both quantitatively and qualitatively.