Youjia Fu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Youjia Fu has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
- Huixia Xue
College of Computer Science and Engineering, Chongqing University of Technology, Chongqing 400054, China
3 shared publications - Junsong Fu
College of Computer Science and Engineering, Chongqing University of Technology, Chongqing 400054, China
3 shared publications - Zihao Xu
Chongqing University of Technology, Chongqing, China
3 shared publications - Antao Lin
School of Computer Science and Engineering, Chongqing University of Technology, Chongqing 400054, China
1 shared publication Huixia Xue
College of Computer Science and Engineering, Chongqing University of Technology, Chongqing, China
1 shared publicationJunsong Fu
College of Computer Science and Engineering, Chongqing University of Technology, Chongqing, China
1 shared publication
Latest Publications
- Open Access
RD-RE: Reverse Distillation with Feature Reconstruction Enhancement for Industrial Anomaly Detection
Industrial anomaly detection methods based on reverse distillation (RD) have shown significant potential. However, existing RD approaches struggle to achieve an effective balance between constraining the feature consistency of the teacher–student networks and maintaining differentiated representation capability, which is crucial for precise anomaly detection. To address this challenge, we propose Reverse Distillation with Feature Reconstruction Enhancement (RD-RE) for Industrial Anomaly Detection. Firstly, we design a cross-stage feature fusion student network to integrate spatial detail information from the encoder with rich semantic information from the decoder. Secondly, we introduce a Locally Aware Dynamic Attention (LDA) module to enhance local detail feature response, thereby improving the model’s robustness in capturing anomalous regions. Finally, a Context-Aware Adaptive Multi-Scale Feature Fusion (CFFMS-FF) module is designed to constrain the consistency of local feature reconstruction. Experiments on the MVTec AD benchmark dataset demonstrate the effectiveness of RD-RE, achieving competitive results of 99.0%, 95.8%, 78.3%, and 99.7% on pixel-level AUROC, PRO, and AP and image-level AUROC metrics, and outperforming existing RD-based approaches. These results conclude that the integration of cross-stage fusion and local attention effectively mitigates the representation-consistency trade-off, providing a more robust solution for industrial anomaly localization.
Open AccessRD-RE: Reverse Distillation with Feature Reconstruction Enhancement for Industrial Anomaly Detection
Industrial anomaly detection methods based on reverse distillation (RD) have shown significant potential. However, existing RD approaches struggle to achieve an effective balance between constraining the feature consistency of the teacher–student networks and maintaining differentiated representation capability, which is crucial for precise anomaly detection. To address this challenge, we propose Reverse Distillation with Feature Reconstruction Enhancement (RD-RE) for Industrial Anomaly Detection. Firstly, we design a cross-stage feature fusion student network to integrate spatial detail information from the encoder with rich semantic information from the decoder. Secondly, we introduce a Locally Aware Dynamic Attention (LDA) module to enhance local detail feature response, thereby improving the model’s robustness in capturing anomalous regions. Finally, a Context-Aware Adaptive Multi-Scale Feature Fusion (CFFMS-FF) module is designed to constrain the consistency of local feature reconstruction. Experiments on the MVTec AD benchmark dataset demonstrate the effectiveness of RD-RE, achieving competitive results of 99.0%, 95.8%, 78.3%, and 99.7% on pixel-level AUROC, PRO, and AP and image-level AUROC metrics, and outperforming existing RD-based approaches. These results conclude that the integration of cross-stage fusion and local attention effectively mitigates the representation-consistency trade-off, providing a more robust solution for industrial anomaly localization.
- 31 January 2025
- Published by Springer Nature in The Journal of Supercomputing
- Vol. 81 (3) , 1-23
Monomm: a multi-scale mamba-enhanced network for real-time monocular 3D object detection
- 31 January 2025
- Published by Springer Nature in The Journal of Supercomputing
- Vol. 81 (3) , 1-23
Monomm: a multi-scale mamba-enhanced network for real-time monocular 3D object detection
- Open Access
- 7 January 2025
- Published by Springer Nature in Complex & Intelligent Systems
- Vol. 11 (2) , 1-16
TMFN: a text-based multimodal fusion network with multi-scale feature extraction and unsupervised contrastive learning for multimodal sentiment analysis
- 7 January 2025
- Published by Springer Nature in Complex & Intelligent Systems
- Vol. 11 (2) , 1-16
Open AccessTMFN: a text-based multimodal fusion network with multi-scale feature extraction and unsupervised contrastive learning for multimodal sentiment analysis
- Open Access
- 18 July 2024
- Published by MDPI AG in Electronics
- Vol. 13 (14) , 2835
Self-HCL: Self-Supervised Multitask Learning with Hybrid Contrastive Learning Strategy for Multimodal Sentiment Analysis
Multimodal Sentiment Analysis (MSA) plays a critical role in many applications, including customer service, personal assistants, and video understanding. Currently, the majority of research on MSA is focused on the development of multimodal representations, largely owing to the scarcity of unimodal annotations in MSA benchmark datasets. However, the sole reliance on multimodal representations to train models results in suboptimal performance due to the insufficient learning of each unimodal representation. To this end, we propose Self-HCL, which initially optimizes the unimodal features extracted from a pretrained model through the Unimodal Feature Enhancement Module (UFEM), and then uses these optimized features to jointly train multimodal and unimodal tasks. Furthermore, we employ a Hybrid Contrastive Learning (HCL) strategy to facilitate the learned representation of multimodal data, enhance the representation ability of multimodal fusion through unsupervised contrastive learning, and improve the model’s performance in the absence of unimodal annotations through supervised contrastive learning. Finally, based on the characteristics of unsupervised contrastive learning, we propose a new Unimodal Label Generation Module (ULGM) that can stably generate unimodal labels in a short training period. Extensive experiments on the benchmark datasets CMU-MOSI and CMU-MOSEI demonstrate that our model outperforms state-of-the-art methods.
- 18 July 2024
- Published by MDPI AG in Electronics
- Vol. 13 (14) , 2835
Open AccessSelf-HCL: Self-Supervised Multitask Learning with Hybrid Contrastive Learning Strategy for Multimodal Sentiment Analysis
Multimodal Sentiment Analysis (MSA) plays a critical role in many applications, including customer service, personal assistants, and video understanding. Currently, the majority of research on MSA is focused on the development of multimodal representations, largely owing to the scarcity of unimodal annotations in MSA benchmark datasets. However, the sole reliance on multimodal representations to train models results in suboptimal performance due to the insufficient learning of each unimodal representation. To this end, we propose Self-HCL, which initially optimizes the unimodal features extracted from a pretrained model through the Unimodal Feature Enhancement Module (UFEM), and then uses these optimized features to jointly train multimodal and unimodal tasks. Furthermore, we employ a Hybrid Contrastive Learning (HCL) strategy to facilitate the learned representation of multimodal data, enhance the representation ability of multimodal fusion through unsupervised contrastive learning, and improve the model’s performance in the absence of unimodal annotations through supervised contrastive learning. Finally, based on the characteristics of unsupervised contrastive learning, we propose a new Unimodal Label Generation Module (ULGM) that can stably generate unimodal labels in a short training period. Extensive experiments on the benchmark datasets CMU-MOSI and CMU-MOSEI demonstrate that our model outperforms state-of-the-art methods.
- Open Access
- 25 June 2024
- Published by MDPI AG in Electronics
- Vol. 13 (13) , 2477
E-DNet: An End-to-End Dual-Branch Network for Driver Steering Intention Detection
An advanced driving assistant system (ADAS) is critical for improving traffic efficiency and ensuring driving safety. By anticipating the driver’s steering intentions in advance, the system can alert the driver in time to avoid a vehicle collision. This paper proposes a novel end-to-end dual-branch network (EDNet) that utilizes both in-cabin and out-of-cabin data. In this study, we designed an in-cabin driver intent feature extractor based on 3D residual networks and atrous convolution, which is applicable to video data and is capable of capturing a larger range of driver behavior. In order to capture the long-term dependency of temporal data, we designed the depthwise-separable max-pooling (DSMax) module and combined it with a convolutional LSTM to obtain the road environment feature extractor outside the cabin. In addition, to effectively fuse different features inside and outside the cockpit, we designed and propose the dynamic combined-feature attention fusion (D-CAF) module. EDNet employs a freeze-training method, which enables the creation of a lightweight model while simultaneously enhancing the final classification accuracy. Extensive experiments on the Brain4Cars dataset and the Zenodo dataset show that the proposed EDNet was able to recognize the driver’s steering intention up to 3 s in advance. It outperformed the existing state of the art in most driving scenarios.
- 25 June 2024
- Published by MDPI AG in Electronics
- Vol. 13 (13) , 2477
Open AccessE-DNet: An End-to-End Dual-Branch Network for Driver Steering Intention Detection
An advanced driving assistant system (ADAS) is critical for improving traffic efficiency and ensuring driving safety. By anticipating the driver’s steering intentions in advance, the system can alert the driver in time to avoid a vehicle collision. This paper proposes a novel end-to-end dual-branch network (EDNet) that utilizes both in-cabin and out-of-cabin data. In this study, we designed an in-cabin driver intent feature extractor based on 3D residual networks and atrous convolution, which is applicable to video data and is capable of capturing a larger range of driver behavior. In order to capture the long-term dependency of temporal data, we designed the depthwise-separable max-pooling (DSMax) module and combined it with a convolutional LSTM to obtain the road environment feature extractor outside the cabin. In addition, to effectively fuse different features inside and outside the cockpit, we designed and propose the dynamic combined-feature attention fusion (D-CAF) module. EDNet employs a freeze-training method, which enables the creation of a lightweight model while simultaneously enhancing the final classification accuracy. Extensive experiments on the Brain4Cars dataset and the Zenodo dataset show that the proposed EDNet was able to recognize the driver’s steering intention up to 3 s in advance. It outperformed the existing state of the art in most driving scenarios.