Di Chen
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
Di Chen has a total of 26 co-authors. Here are the top co-authors by number of shared publications:
- Shanshan Zhang
Nanjing University of Science and Technology
5 shared publications Jian Yang
Nanjing University of Science and Technology
4 shared publications- Bernt Schiele
Max Planck Institute for Informatics, Saarbrücken, Germany
3 shared publications - Jian Yang
Nanjing University of Science and Technology, Nanjing, China
3 shared publications - Yanling Tian
Nanjing University of Science and Technology
2 shared publications - Yunan Liu
Computational Intelligence Center (CIC), School of Computer and Artificial Intelligence, Shandong Jianzhu University, Jinan, China
2 shared publications
Latest Publications
- Open Access
- 24 March 2024
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 38 (6) , 5224-5232
Divide and Conquer: Hybrid Pre-training for Person Search
Large-scale pre-training has proven to be an effective method for improving performance across different tasks. Current person search methods use ImageNet pre-trained models for feature extraction, yet it is not an optimal solution due to the gap between the pre-training task and person search task (as a downstream task). Therefore, in this paper, we focus on pre-training for person search, which involves detecting and re-identifying individuals simultaneously. Although labeled data for person search is scarce, datasets for two sub-tasks person detection and re-identification are relatively abundant. To this end, we propose a hybrid pre-training framework specifically designed for person search using sub-task data only. It consists of a hybrid learning paradigm that handles data with different kinds of supervisions, and an intra-task alignment module that alleviates domain discrepancy under limited resources. To the best of our knowledge, this is the first work that investigates how to support full-task pre-training using sub-task data. Extensive experiments demonstrate that our pre-trained model can achieve significant improvements across diverse protocols, such as person search method, fine-tuning data, pre-training data and model backbone. For example, our model improves ResNet50 based NAE by 10.3% relative improvement w.r.t. mAP. Our code and pre-trained models are released for plug-and-play usage to the person search community (https://github.com/personsearch/PretrainPS).
- 24 March 2024
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 38 (6) , 5224-5232
Open AccessDivide and Conquer: Hybrid Pre-training for Person Search
Large-scale pre-training has proven to be an effective method for improving performance across different tasks. Current person search methods use ImageNet pre-trained models for feature extraction, yet it is not an optimal solution due to the gap between the pre-training task and person search task (as a downstream task). Therefore, in this paper, we focus on pre-training for person search, which involves detecting and re-identifying individuals simultaneously. Although labeled data for person search is scarce, datasets for two sub-tasks person detection and re-identification are relatively abundant. To this end, we propose a hybrid pre-training framework specifically designed for person search using sub-task data only. It consists of a hybrid learning paradigm that handles data with different kinds of supervisions, and an intra-task alignment module that alleviates domain discrepancy under limited resources. To the best of our knowledge, this is the first work that investigates how to support full-task pre-training using sub-task data. Extensive experiments demonstrate that our pre-trained model can achieve significant improvements across diverse protocols, such as person search method, fine-tuning data, pre-training data and model backbone. For example, our model improves ResNet50 based NAE by 10.3% relative improvement w.r.t. mAP. Our code and pre-trained models are released for plug-and-play usage to the person search community (https://github.com/personsearch/PretrainPS).
- Open Access
Divide and Conquer: Hybrid Pre-training for Person Search
Large-scale pre-training has proven to be an effective method for improving performance across different tasks. Current person search methods use ImageNet pre-trained models for feature extraction, yet it is not an optimal solution due to the gap between the pre-training task and person search task (as a downstream task). Therefore, in this paper, we focus on pre-training for person search, which involves detecting and re-identifying individuals simultaneously. Although labeled data for person search is scarce, datasets for two sub-tasks person detection and re-identification are relatively abundant. To this end, we propose a hybrid pre-training framework specifically designed for person search using sub-task data only. It consists of a hybrid learning paradigm that handles data with different kinds of supervisions, and an intra-task alignment module that alleviates domain discrepancy under limited resources. To the best of our knowledge, this is the first work that investigates how to support full-task pre-training using sub-task data. Extensive experiments demonstrate that our pre-trained model can achieve significant improvements across diverse protocols, such as person search method, fine-tuning data, pre-training data and model backbone. For example, our model improves ResNet50 based NAE by 10.3% relative improvement w.r.t. mAP. Our code and pre-trained models are released for plug-and-play usage to the person search community.
Open AccessDivide and Conquer: Hybrid Pre-training for Person Search
Large-scale pre-training has proven to be an effective method for improving performance across different tasks. Current person search methods use ImageNet pre-trained models for feature extraction, yet it is not an optimal solution due to the gap between the pre-training task and person search task (as a downstream task). Therefore, in this paper, we focus on pre-training for person search, which involves detecting and re-identifying individuals simultaneously. Although labeled data for person search is scarce, datasets for two sub-tasks person detection and re-identification are relatively abundant. To this end, we propose a hybrid pre-training framework specifically designed for person search using sub-task data only. It consists of a hybrid learning paradigm that handles data with different kinds of supervisions, and an intra-task alignment module that alleviates domain discrepancy under limited resources. To the best of our knowledge, this is the first work that investigates how to support full-task pre-training using sub-task data. Extensive experiments demonstrate that our pre-trained model can achieve significant improvements across diverse protocols, such as person search method, fine-tuning data, pre-training data and model backbone. For example, our model improves ResNet50 based NAE by 10.3% relative improvement w.r.t. mAP. Our code and pre-trained models are released for plug-and-play usage to the person search community.
- Open Access
Keypoint Message Passing for Video-based Person Re-Identification
Video-based person re-identification (re-ID) is an important technique in visual surveillance systems which aims to match video snippets of people captured by different cameras. Existing methods are mostly based on convolutional neural networks (CNNs), whose building blocks either process local neighbor pixels at a time, or, when 3D convolutions are used to model temporal information, suffer from the misalignment problem caused by person movement. In this paper, we propose to overcome the limitations of normal convolutions with a human-oriented graph method. Specifically, features located at person joint keypoints are extracted and connected as a spatial-temporal graph. These keypoint features are then updated by message passing from their connected nodes with a graph convolutional network (GCN). During training, the GCN can be attached to any CNN-based person re-ID model to assist representation learning on feature maps, whilst it can be dropped after training for better inference speed. Our method brings significant improvements over the CNN-based baseline model on the MARS dataset with generated person keypoints and a newly annotated dataset: PoseTrackReID. It also defines a new state-of-the-art method in terms of top-1 accuracy and mean average precision in comparison to prior works.
Open AccessKeypoint Message Passing for Video-based Person Re-Identification
Video-based person re-identification (re-ID) is an important technique in visual surveillance systems which aims to match video snippets of people captured by different cameras. Existing methods are mostly based on convolutional neural networks (CNNs), whose building blocks either process local neighbor pixels at a time, or, when 3D convolutions are used to model temporal information, suffer from the misalignment problem caused by person movement. In this paper, we propose to overcome the limitations of normal convolutions with a human-oriented graph method. Specifically, features located at person joint keypoints are extracted and connected as a spatial-temporal graph. These keypoint features are then updated by message passing from their connected nodes with a graph convolutional network (GCN). During training, the GCN can be attached to any CNN-based person re-ID model to assist representation learning on feature maps, whilst it can be dropped after training for better inference speed. Our method brings significant improvements over the CNN-based baseline model on the MARS dataset with generated person keypoints and a newly annotated dataset: PoseTrackReID. It also defines a new state-of-the-art method in terms of top-1 accuracy and mean average precision in comparison to prior works.
- 14 September 2021
- Published by Springer Nature in International Journal of Computer Vision
- Vol. 129 (11) , 3154-3168
Norm-Aware Embedding for Efficient Person Search and Tracking
- 14 September 2021
- Published by Springer Nature in International Journal of Computer Vision
- Vol. 129 (11) , 3154-3168
Norm-Aware Embedding for Efficient Person Search and Tracking
- 8 April 2021
- Published by Springer Nature in International Journal of Computer Vision
- Vol. 129 (6) , 1875-1892
Guided Attention in CNNs for Occluded Pedestrian Detection and Re-identification
- 8 April 2021
- Published by Springer Nature in International Journal of Computer Vision
- Vol. 129 (6) , 1875-1892
Guided Attention in CNNs for Occluded Pedestrian Detection and Re-identification
- Open Access
- 4 March 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Proceedings of the IEEE
- Vol. 109 (9) , 1494-1520
Advances in Video Compression System Using Deep Neural Network: A Review and Case Studies
Significant advances in video compression systems have been made in the past several decades to satisfy the near-exponential growth of Internet-scale video traffic. From the application perspective, we have identified three major functional blocks, including preprocessing, coding, and postprocessing, which have been continuously investigated to maximize the end-user quality of experience (QoE) under a limited bit rate budget. Recently, artificial intelligence (AI)-powered techniques have shown great potential to further increase the efficiency of the aforementioned functional blocks, both individually and jointly. In this article, we review recent technical advances in video compression systems extensively, with an emphasis on deep neural network (DNN)-based approaches, and then present three comprehensive case studies. On preprocessing, we show a switchable texture-based video coding example that leverages DNN-based scene understanding to extract semantic areas for the improvement of a subsequent video coder. On coding, we present an end-to-end neural video coding framework that takes advantage of the stacked DNNs to efficiently and compactly code input raw videos via fully data-driven learning. On postprocessing, we demonstrate two neural adaptive filters to, respectively, facilitate the in-loop and postfiltering for the enhancement of compressed frames. Finally, a companion website hosting the contents developed in this work can be accessed publicly at https://purdueviper.github.io/dnn-coding/.
- 4 March 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Proceedings of the IEEE
- Vol. 109 (9) , 1494-1520
Open AccessAdvances in Video Compression System Using Deep Neural Network: A Review and Case Studies
Significant advances in video compression systems have been made in the past several decades to satisfy the near-exponential growth of Internet-scale video traffic. From the application perspective, we have identified three major functional blocks, including preprocessing, coding, and postprocessing, which have been continuously investigated to maximize the end-user quality of experience (QoE) under a limited bit rate budget. Recently, artificial intelligence (AI)-powered techniques have shown great potential to further increase the efficiency of the aforementioned functional blocks, both individually and jointly. In this article, we review recent technical advances in video compression systems extensively, with an emphasis on deep neural network (DNN)-based approaches, and then present three comprehensive case studies. On preprocessing, we show a switchable texture-based video coding example that leverages DNN-based scene understanding to extract semantic areas for the improvement of a subsequent video coder. On coding, we present an end-to-end neural video coding framework that takes advantage of the stacked DNNs to efficiently and compactly code input raw videos via fully data-driven learning. On postprocessing, we demonstrate two neural adaptive filters to, respectively, facilitate the in-loop and postfiltering for the enhancement of compressed frames. Finally, a companion website hosting the contents developed in this work can be accessed publicly at https://purdueviper.github.io/dnn-coding/.
- 19 February 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 29 (10577149) , 4669-4682
Person Search by Separated Modeling and A Mask-Guided Two-Stream CNN Model
In this work, we tackle the problem of person search, which is a challenging task consisted of pedestrian detection and person re-identification (re-ID). Instead of sharing representations in a single joint model, we find that separating detector and re-ID feature extraction yields better performance. In order to extract more representative features for each identity, we segment out the foreground person from the original image patch. We propose a simple yet effective re-ID method, which models foreground person and original image patches individually, and obtains enriched representations from two separate CNN streams. We also propose a Confidence Weighted Stream Attention method which further re-adjusts the relative importance of the two streams by incorporating the detection confidence. Furthermore, we simplify the whole pipeline by incorporating semantic segmentation into the re-ID network, which is trained by bounding boxes as weakly-annotated masks and identification labels simultaneously. From the experiments on two standard person search benchmarks i.e. CUHK-SYSU and PRW, we achieve mAP of 83.3% and 32.8% respectively, surpassing the state of the art by a large margin. The extensive ablation study and model inspection further justifies our motivation.
- 19 February 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 29 (10577149) , 4669-4682
Person Search by Separated Modeling and A Mask-Guided Two-Stream CNN Model
In this work, we tackle the problem of person search, which is a challenging task consisted of pedestrian detection and person re-identification (re-ID). Instead of sharing representations in a single joint model, we find that separating detector and re-ID feature extraction yields better performance. In order to extract more representative features for each identity, we segment out the foreground person from the original image patch. We propose a simple yet effective re-ID method, which models foreground person and original image patches individually, and obtains enriched representations from two separate CNN streams. We also propose a Confidence Weighted Stream Attention method which further re-adjusts the relative importance of the two streams by incorporating the detection confidence. Furthermore, we simplify the whole pipeline by incorporating semantic segmentation into the re-ID network, which is trained by bounding boxes as weakly-annotated masks and identification labels simultaneously. From the experiments on two standard person search benchmarks i.e. CUHK-SYSU and PRW, we achieve mAP of 83.3% and 32.8% respectively, surpassing the state of the art by a large margin. The extensive ablation study and model inspection further justifies our motivation.
- 1 January 2018
- Published by Elsevier in Neurocomputing
- Vol. 275, 560-567
Joint Bayesian guided metric learning for end-to-end face verification
- 1 January 2018
- Published by Elsevier in Neurocomputing
- Vol. 275, 560-567
Joint Bayesian guided metric learning for end-to-end face verification
- Open Access
- 1 January 2018
- Published by Association for Computational Linguistics (ACL)
- p. 665-670
Hybrid Neural Attention for Agreement/Disagreement Inference in Online Debates
- 1 January 2018
- Published by Association for Computational Linguistics (ACL)
- p. 665-670
Open AccessHybrid Neural Attention for Agreement/Disagreement Inference in Online Debates
- 22 March 2015
- Published by SAGE Publications in Textile Research Journal
- Vol. 85 (18) , 1972-1984
Computer simulation of batik printing patterns with cracks
Generating batik printing patterns is an ancient and unique dye-resistant printing technique. Due to the dip dyeing of the cloth with wax in the dyes, the wax as a reserve dyeing agent cracks naturally, and the dyes permeate into the cloth along the cracks. It forms a natural texture and generates a unique and beautiful artwork that is difficult to produce manually. According to the characteristics of the batik printing patterns with cracks, we developed a specific method for simulation of the batik printing patterns with cracks that are relevant or not relevant to the batik patterns (BPPCNR or BPPCR). For the simulation of the process of the extension of the crack, we adapted the method of solving two-dimensional crack fracture based on the extended finite element method (EFEM) in the field of fracture mechanics. We achieved the results of cracks with different shades of colors for batik printing patterns with cracks that are not relevant to the batik patterns (BPPCNR). For the batik printing patterns with cracks that are relevant to the batik patterns (BPPCR), we proposed the method combining the distance transformation algorithm and EFEM, which can quickly define the initial cracks by the distance transformation image of the input printing patterns, and simulated the process of the extension of the crack by the EFEM. Moreover, we proposed the definition of a local axis parallel extension area that can greatly reduce the computation overhead. Experiments show that our method is effective in that it can realize various satisfying vivid batik printing patterns with cracks.
- 22 March 2015
- Published by SAGE Publications in Textile Research Journal
- Vol. 85 (18) , 1972-1984
Computer simulation of batik printing patterns with cracks
Generating batik printing patterns is an ancient and unique dye-resistant printing technique. Due to the dip dyeing of the cloth with wax in the dyes, the wax as a reserve dyeing agent cracks naturally, and the dyes permeate into the cloth along the cracks. It forms a natural texture and generates a unique and beautiful artwork that is difficult to produce manually. According to the characteristics of the batik printing patterns with cracks, we developed a specific method for simulation of the batik printing patterns with cracks that are relevant or not relevant to the batik patterns (BPPCNR or BPPCR). For the simulation of the process of the extension of the crack, we adapted the method of solving two-dimensional crack fracture based on the extended finite element method (EFEM) in the field of fracture mechanics. We achieved the results of cracks with different shades of colors for batik printing patterns with cracks that are not relevant to the batik patterns (BPPCNR). For the batik printing patterns with cracks that are relevant to the batik patterns (BPPCR), we proposed the method combining the distance transformation algorithm and EFEM, which can quickly define the initial cracks by the distance transformation image of the input printing patterns, and simulated the process of the extension of the crack by the EFEM. Moreover, we proposed the definition of a local axis parallel extension area that can greatly reduce the computation overhead. Experiments show that our method is effective in that it can realize various satisfying vivid batik printing patterns with cracks.