A5018182257- Multimodal Machine Learning Applications
- Advanced Neural Network Applications
- Anomaly Detection Techniques and Applications
- Text and Document Classification Technologies
- Domain Adaptation and Few-Shot Learning
- Data-Driven Disease Surveillance
Huazhong University of Science and Technology
2022-2024
Recently, the semantics of scene text has been proven to be essential in fine-grained image classification. However, existing methods mainly exploit literal meaning for recognition, which might irrelevant when it is not significantly related objects/scenes. We propose an end-to-end trainable network that mines implicit contextual knowledge behind and enhance correlation fine-tune representation. Unlike methods, our model integrates three modalities: visual feature extraction, correlating...
<title>Abstract</title> Semantic segmentation is a fundamental problem in computer vision and it requires high-resolution feature maps for dense prediction. Current coordinate-guided low-resolution interpolation methods, e.g., bilinear interpolation, produce coarse features which suffer from misalignment insufficient context information. Moreover, enriching semantics to high computation burden, so that challenging meet the requirement of lowlatency inference. We propose novel Guided...
Recently, the semantics of scene text has been proven to be essential in fine-grained image classification. However, existing methods mainly exploit literal meaning for recognition, which might irrelevant when it is not significantly related objects/scenes. We propose an end-to-end trainable network that mines implicit contextual knowledge behind and enhance correlation fine-tune representation. Unlike methods, our model integrates three modalities: visual feature extraction, correlating...