A5063689776- Video Surveillance and Tracking Methods
- Impact of Light on Environment and Health
- Fire Detection and Safety Systems
- Remote Sensing in Agriculture
- Spectroscopy and Chemometric Analyses
- Remote Sensing and Land Use
- Remote-Sensing Image Classification
- Smart Agriculture and AI
- IoT-based Smart Home Systems
Nanjing University of Information Science and Technology
2021-2023
Shandong Agricultural University
2018-2019
Due to their excellent performance on aggregating global features, Transformer structures are being widely employed in deep learning-based visual object tracking algorithms, recently. Nevertheless, existing Transformer-based trackers still fail handle occlusion problems due drift feature distributions. To address this issue, we introduce domain adaptation techniques into a novel framework, DATransT, including extraction, adaptive module and prediction head. The consists of three...
When the spatial distribution of winter wheat is extracted from high-resolution remote sensing imagery using convolutional neural networks (CNN), field edge results are usually rough, resulting in lowered overall accuracy. This study proposed a new per-pixel classification model CNN and Bayesian models (CNN-Bayesian model) for improved extraction In this model, feature extractor generates vector each pixel, an encoder transforms pixel into category-code vector, two-level classifier uses...
When extracting winter wheat spatial distribution by using convolutional neural network (CNN) from Gaofen-2 (GF-2) remote sensing images, accurate identification of edge pixel is the key to improving result accuracy. In this paper, an approach for based on CNN proposed. A hybrid structure (HSCNN) was first constructed, which consists two independent sub-networks different depths. The deeper sub-network used extract pixels present in interior field, whereas shallower extracts at field. model...
With the rapid development of deep learning techniques, new breakthroughs have been made in learning-based object tracking methods. Although many approaches achieved state-of-the-art results, existing methods still cannot fully satisfy practical needs. A robust tracker should perform well three aspects: accuracy, speed, and resource consumption. Considering this notion, we propose a novel model, Faster MDNet, to strike better balance among these factors. To improve channel attention module...
The correlation filter method is effective in visual tracking tasks, whereas it suffers from the boundary effect and degradation complex situations, which can result suboptimal performance. Aiming at solving above problem, this study proposes an object with a discriminant filter, combines adaptive background perception spatial dynamic constraint. In method, background-awareness strategy used to information trained by interference improve discriminability between background. addition,...
Accuracy and speed have been two fundamental issues that are difficult to balance in object tracking. Trackers with high accuracy often quite large network structures require huge amounts of computing resources, therefore leading a lower tracking speed. To address the problem, we propose novel domain adaptive algorithm obtain better between accuracy. A simple effective adaptation component is employed transfer features from image classification domain. In addition, construct an spatial...