Kenan Lou
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Kenan Lou has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Hao Xu
College of Engineering, Lishui University, Lishui 323000, China
3 shared publications - Yang Shen
School of Artificial Intelligence, Lishui University, Lishui City 323000, China
3 shared publications Ruihua Zhang
College of Engineering, Lishui University, Lishui, China
2 shared publicationsShuai Wang
CNRS, LPL, Aix Marseille University, Aix-en-Provence, France
2 shared publications- Shuyue Zhou
School of Informatics, Xiamen University
2 shared publications Hao. Xu
Department of Computer Science and Technology, Lishui University, Lishui, China
1 shared publication
Latest Publications
- Open Access
IHPP: Improved Human Parts as Points for Multi-Person Pose Regression
Multi-personpose estimation is a fundamental technology for deep learning-based intelligent sensing systems, enabling downstream understanding of human action in applications such as surveillance, robotics, healthcare, and sports analytics. Most two-stage multi-person pose estimation algorithms suffer from low efficiency due to their decoupled representation between body and keypoints, resulting in a complex inference process. Single-stage algorithms such as AdaptivePose that represent body parts as semantic proxy points can substantially simplify the pipeline while achieving competitive results; however, the limited receptive field of convolutional features still makes it difficult to localize proxy points for long-range extremities in semantically informative regions. IHPP addresses this issue as a task-specific refinement of AdaptivePose; it enriches the proxy-point perceiver with directional context modeling along the horizontal and vertical axes and redesigns the second-step regression stage with a part-wise branch to reduce inter-part feature interference while keeping the overall pipeline lightweight. Here, we present IHPP in three sizes (IHPP-S, IHPP-M, and IHPP-L), optimizing the tradeoffs between efficiency and accuracy through fine-tuning of the channel dimensions for each bodypart feature. IHPP-L achieves 72.3 AP at 28 fps on COCO test-dev, surpassing DEKR-W48 and SWAHR-W48 by 1.3 AP and 0.3 AP, respectively, while IHPP-M reaches 69.0 AP at 35 fps and IHPP-S runs at 42 fps with only 9.4 M parameters. On CrowdPose, IHPP-M outperforms AdaptivePose-W48 by 0.2 AP with about one-third the parameters. Comprehensive experiments on the MS COCO and CrowdPose datasets validate the effectiveness of this design.
Open AccessIHPP: Improved Human Parts as Points for Multi-Person Pose Regression
Multi-personpose estimation is a fundamental technology for deep learning-based intelligent sensing systems, enabling downstream understanding of human action in applications such as surveillance, robotics, healthcare, and sports analytics. Most two-stage multi-person pose estimation algorithms suffer from low efficiency due to their decoupled representation between body and keypoints, resulting in a complex inference process. Single-stage algorithms such as AdaptivePose that represent body parts as semantic proxy points can substantially simplify the pipeline while achieving competitive results; however, the limited receptive field of convolutional features still makes it difficult to localize proxy points for long-range extremities in semantically informative regions. IHPP addresses this issue as a task-specific refinement of AdaptivePose; it enriches the proxy-point perceiver with directional context modeling along the horizontal and vertical axes and redesigns the second-step regression stage with a part-wise branch to reduce inter-part feature interference while keeping the overall pipeline lightweight. Here, we present IHPP in three sizes (IHPP-S, IHPP-M, and IHPP-L), optimizing the tradeoffs between efficiency and accuracy through fine-tuning of the channel dimensions for each bodypart feature. IHPP-L achieves 72.3 AP at 28 fps on COCO test-dev, surpassing DEKR-W48 and SWAHR-W48 by 1.3 AP and 0.3 AP, respectively, while IHPP-M reaches 69.0 AP at 35 fps and IHPP-S runs at 42 fps with only 9.4 M parameters. On CrowdPose, IHPP-M outperforms AdaptivePose-W48 by 0.2 AP with about one-third the parameters. Comprehensive experiments on the MS COCO and CrowdPose datasets validate the effectiveness of this design.
- Open Access
- 11 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 54066-54072
Geometric Pooling: Maintaining More Representative Information
Graph Pooling technology plays an important role in graph node classification tasks. Sorting pooling technologies maintain large-value units for pooling graphs of varying sizes. However, by analyzing the statistical characteristic of activated units after pooling, we found that a large number of units dropped by sorting pooling are negative-value units that contain representative information and can contribute considerably to the final decision. To maintain more representative information, we proposed a novel pooling technology, called Geometric Pooling (GP), containing the unique node features with negative values by measuring the similarity of all node features. We reveal the effectiveness of GP from the entropy reduction view. The experiments were conducted on TUdatasets to show the effectiveness of GP. The results showed that the proposed GP outperforms the SOTA graph pooling technologies by $1\%~\sim ~5\%$ with fewer parameters.
- 11 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 54066-54072
Open AccessGeometric Pooling: Maintaining More Representative Information
Graph Pooling technology plays an important role in graph node classification tasks. Sorting pooling technologies maintain large-value units for pooling graphs of varying sizes. However, by analyzing the statistical characteristic of activated units after pooling, we found that a large number of units dropped by sorting pooling are negative-value units that contain representative information and can contribute considerably to the final decision. To maintain more representative information, we proposed a novel pooling technology, called Geometric Pooling (GP), containing the unique node features with negative values by measuring the similarity of all node features. We reveal the effectiveness of GP from the entropy reduction view. The experiments were conducted on TUdatasets to show the effectiveness of GP. The results showed that the proposed GP outperforms the SOTA graph pooling technologies by $1\%~\sim ~5\%$ with fewer parameters.
- Open Access
- 13 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 27017-27028
Style-Based Tree GAN for Point Cloud Generator
Point cloud generation and representation is important in industry areas. Generating and editing high quality 3D shapes is challenging work in deep learning. Inspired by StyleGAN, a style based generative adversarial networks is proposed to generate high quality 3D point cloud. An improved non-linear mapping network learn distribution of points and is used to generate well distributed point cloud point cloud. We also provide a coarse to fine representation for point cloud. According to the experimental results on the ShapeNet Part data set(including aircraft, chair single category and overall 16 categories), our method can generate more uniform point cloud than other GAN methods with less training epoches. The latent code for point cloud has better linear separation,and is more easy to edit.
- 13 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 27017-27028
Open AccessStyle-Based Tree GAN for Point Cloud Generator
Point cloud generation and representation is important in industry areas. Generating and editing high quality 3D shapes is challenging work in deep learning. Inspired by StyleGAN, a style based generative adversarial networks is proposed to generate high quality 3D point cloud. An improved non-linear mapping network learn distribution of points and is used to generate well distributed point cloud point cloud. We also provide a coarse to fine representation for point cloud. According to the experimental results on the ShapeNet Part data set(including aircraft, chair single category and overall 16 categories), our method can generate more uniform point cloud than other GAN methods with less training epoches. The latent code for point cloud has better linear separation,and is more easy to edit.
- Open Access
- 14 December 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 167076-167083
Kernel Product Neural Networks
Attention is an important field to explore the importance of each convolutional kernel channel/weight. The existing attention methods mostly use the Squeeze-and-Excitation (SE) technology to extract the global nonlinear feature vectors as the weights of corresponding feature maps. However, the pooling operators and fully-connected layers used in SE technology extract global features at the cost of much valuable information loss and the parameter amount increase. Actually, the feature map containing full information is a ready-made and better attention for other feature maps in the same layer. Simultaneously the products of feature maps will bring powerful non-linearity. Seeing this, Kernel Product (KP) technology is proposed to simply get useful nonlinear attention. To verify the effectiveness of KP, the proposed KP module is employed on Selective Kernel Networks (SKNets) to take the place of the original SE technology. The variety of SKNets is called Kernel Product Networks (KPNets) in this paper. In addition, identity mapping is used to solve the non-convergence problem in very deep neural networks. The KPNets are evaluated on ImageNet-1k, CIFAR-10, and CIFAR-100. The experiment results show that KPNets outperform many state-of-the-art methods and get a similar but more efficient performance than its SKNets with counterpart.
- 14 December 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 167076-167083
Open AccessKernel Product Neural Networks
Attention is an important field to explore the importance of each convolutional kernel channel/weight. The existing attention methods mostly use the Squeeze-and-Excitation (SE) technology to extract the global nonlinear feature vectors as the weights of corresponding feature maps. However, the pooling operators and fully-connected layers used in SE technology extract global features at the cost of much valuable information loss and the parameter amount increase. Actually, the feature map containing full information is a ready-made and better attention for other feature maps in the same layer. Simultaneously the products of feature maps will bring powerful non-linearity. Seeing this, Kernel Product (KP) technology is proposed to simply get useful nonlinear attention. To verify the effectiveness of KP, the proposed KP module is employed on Selective Kernel Networks (SKNets) to take the place of the original SE technology. The variety of SKNets is called Kernel Product Networks (KPNets) in this paper. In addition, identity mapping is used to solve the non-convergence problem in very deep neural networks. The KPNets are evaluated on ImageNet-1k, CIFAR-10, and CIFAR-100. The experiment results show that KPNets outperform many state-of-the-art methods and get a similar but more efficient performance than its SKNets with counterpart.