Several studies suggest that sleep quality is associated with physical activities. Moreover, deep sleep time can be used to determine the sleep quality of an individual. In this work, we aim to find the association be...
详细信息
Several studies suggest that sleep quality is associated with physical activities. Moreover, deep sleep time can be used to determine the sleep quality of an individual. In this work, we aim to find the association between physical activities and deep sleep time by modeling the time series data such as heart rate and a number of steps captured from a commercial wearable device. Our previous study demonstrates that deep learning-based time series modeling is well suited for our problem since the temporal patterns in the two physical parameters need to be captured to obtain more accurate results. We first preprocess our series data to have a time-step size of 10 minutes. To improve our previous effort in this modeling, we compare four different variants of Long Short-Term Memory (LSTM)-based models, ranging from single input to dual input models. Our result shows that the simple stacked LSTM model performs better for our data because the remaining models suffer from overfitting due to a larger number of the trained parameters.
Existing music-driven 3D dance generation methods mainly concentrate on high-quality dance generation, but lack sufficient control during the generation process. To address these issues, we propose a unified framework...
Existing music-driven 3D dance generation methods mainly concentrate on high-quality dance generation, but lack sufficient control during the generation process. To address these issues, we propose a unified framework capable of generating high-quality dance movements and supporting multi-modal control, including genre control, semantic control, and spatial control. First, we decouple the dance generation network from the dance control network, thereby avoiding the degradation in dance quality when adding additional control information. Second, we design specific control strategies for different control information and integrate them into a unified framework. Experimental results show that the proposed dance generation framework outperforms state-of-the-art methods in terms of motion quality and controllability.
Facial expression recognition is a hot research topic in artificial intelligence industry and has a good research prospect in various fields. At present, facial expression recognition processes the whole face directly...
详细信息
Convolutional Neural network is state of the art of image recognition or image classification. However to build the robust model using CNN needs many parameters adjusted, and choosing the good combination hyperparamet...
详细信息
Wireless sensor networks are widely valued for their effectiveness in real-time data collection. As the amount of data exchanged within such networks grows, designing a robust network topology that maximizes area cove...
Wireless sensor networks are widely valued for their effectiveness in real-time data collection. As the amount of data exchanged within such networks grows, designing a robust network topology that maximizes area coverage with minimal sensors has become a critical challenge. The choice of topology impacts key network metrics, including sensor coverage, communication range, connectivity, inference, and installation and management costsIn this paper, we address the Wireless Sensor Network Planning Problem with Multiple Sources/Destinations, presenting an optimization approach based on deep reinforcement learning. This problem is noteworthy, as sensors in various applications are often required to share data within distinct destinationsWe leverage deep reinforcement learning to effectively address the complex task of selecting optimal sensor locations. Our reinforcement learning agent dynamically learns network structure by iteratively adding and removing sensors, optimizing both sensor coverage and the total number of sensors used. Experiment across diverse scenarios demonstrate the effectiveness of our method for network planning problems of varying scales, achieving full coverage with fewer sensors than traditional approaches. Additionally, our approach also produce solutions for large instances where Mixed Integer programming solvers were not able to. Overall, our method was able to reduce the number of sensors used by up to 22.3% compared to other methods.
Graph Neural Network (GNN) resembles the diffusion process, leading to the oversmoothing of learned representations when stacking many layers. Hence, the reverse process of message passing can produce the distinguisha...
Graph Neural Network (GNN) resembles the diffusion process, leading to the oversmoothing of learned representations when stacking many layers. Hence, the reverse process of message passing can produce the distinguishable node representations by inverting the forward message propagation. The distinguishable representations can help us to better classify neighboring nodes with different labels, such as in heterophilic graphs. In this work, we apply the design principle of the reverse process to the three variants of the GNNs. Through the experiments on heterophilic graph data, where adjacent nodes need to have different representations for successful classification, we show that the reverse process significantly improves the prediction performance in many cases. Additional analysis reveals that the reverse mechanism can mitigate the over-smoothing over hundreds of layers. Our code is available at https://***/ml-postech/reverse-gnn.
The protection of intellectual property rights (IPR) has taken on increased importance in today’s interconnected world, but it also faces new obstacles. As patents, copyrights, trademarks, and trade secrets are incre...
详细信息
With the recent development of information technology, the importance of protecting personal information has increased. Because of the vulnerability in passwords, biometric authentication is now being used as a method...
详细信息
ISBN:
(数字)9798350378214
ISBN:
(纸本)9798350378221
With the recent development of information technology, the importance of protecting personal information has increased. Because of the vulnerability in passwords, biometric authentication is now being used as a method of personal information protection. However, biometric authentication has the possibility of malicious authentication due to the emergence of technology that can generate biometric data that is difficult to duplicate. Therefore, this study aimed to solve the security problems posed by Generative Adversarial Networks (GANs) in biometric authentication by using Hyperspectral Images (HSI). We searched for solutions to the problems identified in previous studies, such as the inability to correctly identify individuals who have not been previously learned and the length of time required for identification. In this paper, we used three channels of data randomly selected from hyperspectral face images for learning to identify individuals by binary classification into “learned subject (class)” and “unlearned subjects (classes)”. In this case, discriminators based on the least-squares generative adversarial network (LSGAN) model was developed for each class using three randomly selected channels of data from the HSI for two class classification. The proposed method significantly reduced the identification time and achieved very high discrimination accuracy in all classes. In addition, the results showed high accuracy throughout multiple face identification experiments.
In this study, two deep learning models for automatic tattoo detection were analyzed; a modified Convolutional Neural Network (CNN) and pre-trained ResNet-50 model. In order to achieve this, ResNet-50 uses transfer le...
详细信息
ISBN:
(数字)9798350364538
ISBN:
(纸本)9798350364545
In this study, two deep learning models for automatic tattoo detection were analyzed; a modified Convolutional Neural Network (CNN) and pre-trained ResNet-50 model. In order to achieve this, ResNet-50 uses transfer learning with fine-tuning. The purpose of this study was to evaluate the accuracy, precision, recall, F1-score, and computational efficiency of the system being considered. To augment the dataset included 1000 photos that were equally divided between those showing tattoos and those that did not show tattoos. A k-fold cross-validation approach was employed in training and testing the models. Although custom CNNs are effective, utilizing pre-trained ones like ResNet-50 can offer even better outcomes. Specifically, ResNet-50 attained a higher accuracy (0.86 compared to 0.79), precision (0.85 versus 0.78), recall (0.91 against 0.86), and F1-score (0.91 vis-a-vis 0.86) as compared to custom CNNs. In selecting these models for examination, two main motivations were considered. The first motivation is to see whether transfer learning with a pre-trained ResNet-50 model does well when compared with a customized CNN designed specifically for tattoo detection. Secondly,the intent of this study is to know what advantages can be derived from each approach and their demerits too. Furthermore, it seeks to determine if transfer learning can provide an alternative in contrast to the common CNN techniques with regards to precision and computational efficiency. In this research, two models will be evaluated in order to answer the question of what is better for tattoo detection: transfer learning or designing custom architectures.
Inevery nation, the main goal of a presidential election is to choose a good leader. Citizens of a country need to thoroughly observe and assess the presidential presidential candidates before making their selection. ...
详细信息
暂无评论