Previous works on unsupervised skeleton-based action recognition primarily focused on strategies for utilizing features to drive model optimization through methods like contrastive learning and reconstruction. However...
详细信息
Previous works on unsupervised skeleton-based action recognition primarily focused on strategies for utilizing features to drive model optimization through methods like contrastive learning and reconstruction. However, designing application-level strategies poses challenges. This paper shifts the focus to the generation-level modelings and introduces the Spatiotemporal Adaptively Attentions-guided Refining Network (AgRNet). AgRNet approaches the reduction of costs and enhancement of efficiency by constructing the Adaptive Activity- Guided Attention (AAGA) and Adaptive Dominant-Guided Attenuation (ADGA) modules. The AAGA leverages the sparsity of the correlation matrix in the attention mechanism to adaptively filter and retain the active components of the sequence during the modeling process. The ADGA embeds the local dominant features of the sequence, obtained through convolutional distillation, into the globally dominant features under the attention mechanism, guided by the defined attenuation factor. Additionally, the Progressive Feature Modeling (PFM) module is introduced to complement the progressive features in motion sequences that were overlooked by AAGA and ADGA. AgRNet shows efficiency on three public datasets, NTU-RGBD 60, NTU-RGBD 120, and UWA3D. IEEE
The Cloud IoT paradigm, designed to combine Cloud Computing (CC) and the Internet of Things (IoT) benefits, is increasingly utilized for extensive services and addressing users' connectivity, data processing, and ...
详细信息
The growing usage of security cameras in smart cities to enable round-the-clock surveillance has allowed researchers to analyze a vast amount of data. A better security system is required in other monitoring industrie...
详细信息
We present a new xor-based attention function for efficient hardware implementation of transformers. While the standard attention mechanism relies on matrix multiplication between the key and the transpose of the quer...
详细信息
Different product characteristics and consumer expectations must be analyzed when making a product or service recommendation based on use. However, if all types of knowledge are inaccessible, this is known as the clod...
详细信息
The lack of available smoke detectors outdoors constitutes a significant factor contributing to the rapid spread of fires before effective containment can be achieved. The integration of outdoor smoke detectors helps ...
详细信息
The surge in digital transactions has paved the way for an alarming rise in credit card fraud, compelling the need for robust detection systems. The swift progress of technology has transformed customer payment habits...
详细信息
In the recent past, cancer is designated as the life threatening disease causing significant deaths across the globe. Countries are putting efforts for early stage cancer detection, and mortality prediction. Machine L...
详细信息
Lung cancer is a dangerous disease with differing treatment plans based on types and location of the cancerous cells. The overall 5-year survival rate for all stages of lung cancer is around 15%. People who smoke are ...
详细信息
Lung cancer is a dangerous disease with differing treatment plans based on types and location of the cancerous cells. The overall 5-year survival rate for all stages of lung cancer is around 15%. People who smoke are at the highest risk of developing lung cancer. Early detection of lung cancer is crucial for starting early treatment and preventing the disease from spreading. Hence, it can improve people’s chances of survival. Imaging tests, such as a chest computed tomography (CT) scan, can detect lung cancer by providing a more detailed picture. However, the examination of chest CT scans is a challenging task and is prone to subject variability. For this, researchers have developed many computer-aided diagnostic (CAD) systems for the automatic detection of cancer using CT scan images. Misdiagnoses can occur in manual interpretation of images. An automated trained neural network on lung images from healthy and malignant lung cells helps lower the problem. Convolutional neural network (CNN)-based pretrained deep learning models have been used successfully to detect lung cancer. The accuracy of classification is significant to avoid false prediction. This research presents a metalearning based approach for identifying the common types of lung cancer tissues namely, Benign tissue, Squamous Cell Carcinoma, and Adenocarcinoma using LC25000 dataset. All the experiments have been conducted on a publicly available benchmark dataset for lung histopathological images. The features extracted from the penultimate layer (global average pooling) of the transfer learning-based CNN models, namely InceptionResNetV1, EfficientNetB7, and DenseNet121, have been fused together, and the dimensionality reduction has been applied to them before passing to the metaclassifier, which is the Support Vector Machine (SVM) classifier in our case. A quantitative analysis of the proposed algorithm has been conducted through classification accuracy and confusion matrix computation. When compared wit
In the current era, technology helps a lot to enhance human experiences leading to smart city development, which includes optimized energy usage, better healthcare availability, minimal governance, and intelligent mob...
详细信息
暂无评论