A5037688515- Advanced Neural Network Applications
- Advanced Image and Video Retrieval Techniques
- Domain Adaptation and Few-Shot Learning
- Multimodal Machine Learning Applications
- Currency Recognition and Detection
- Medical Image Segmentation Techniques
- Video Surveillance and Tracking Methods
- Visual Attention and Saliency Detection
- Industrial Vision Systems and Defect Detection
Henan University of Science and Technology
2023-2024
Abstract The traditional complete dual-branch structure is effective for semantic segmentation tasks. However, it redundant in some sense. Moreover, the simple additive fusion of features from two branches may not achieve satisfactory performance. To alleviate these problems, this paper we propose an efficient compact interactive network (CIDNet) real-time segmentation. Specifically, first build a by constructing detail branch and branch. Furthermore, detail-semantic module to fuse several...
Abstract Semantic segmentation is a fundamental technology for autonomous driving. It has high demand inference speed and accuracy. However, good trade‐off between accuracy latency yet not present in existing semantic approaches. Due to the limitation of speed, authors cannot increase number network layers without limit design modules like networks real‐time. challenging problem how model with performance under limited resources. To alleviate these issues, this study, propose refinement...
With the development of image segmentation technology, context information plays an increasingly important role in semantic segmentation. However, due to complexity different feature maps, simple capture operations can easily cause omission. Rich better classify categories and improve quality On contrary, poor will lead blurred category incomplete target edge. In order rich as completely possible, we constructed a Multi-Pooling Context Network (MPCNet), which is multi-pool contextual network...
Numerous deep learning-based object detection methods have achieved excellent performance. However, the performance on small-size and positive negative sample imbalance problems is not satisfactory. We propose a multi-scale feature selective matching network (MFSMNet) to improve of alleviate problems. First, we construct semantic enhancement module (MSEM) compensate for information loss small-sized targets during down-sampling by obtaining richer from features at multiple scales. Then,...
Abstract For the purpose of object detection, numerous key points based methods have been suggested. To alleviate imbalance problem that some objects may be missing when a single-center-point network is used for we propose brand-new multiple space cascaded center point (MSCCPNet) detection. Particularly, first bulid novel structure to in detecting different scale by scanning more spaces. We then predict category and confidence integrating results two centers with idea choosing high...
Abstract The semantic information can ensure better pixel classification, and the spatial of low-level feature map detailed location pixels. However, this part is often ignored in capturing information, it a huge loss for image category itself. To alleviate problem, we propose Long Short-Range Relevance Context Network. Specifically, first construct Long-Range Module to capture global context high-level local information. At same time, build piecewise each stage features form jump...