EAI Endorsed Transactions on e-Learning

A fast image inpainting algorithm based on an adaptive scanning strategy

H. R. Guo, W. H. Wang — Tue, 15 Aug 2023 00:00:00 +0000

OBJECTIVES: In exemplar-based image inpainting algorithms, there are often issues with the calculation of patch similarity for matching, suboptimal strategies for selecting matching patches, and low inpainting speed.

METHODS: This paper first uses the variable scale cross-scan block line progressive scan to solve the problem of slow scanning speed and invalid priority formula. Then, an improved weight similarity formula is used for searching to solve the problem of poor computing strategy for similar matching patches. The search range of matching patches gradually increases from small to large until globally searching for similar matching patches to improve the efficiency of inpainting. To further improve the correctness of matching patch selection, this paper uses six levels of priority matching criteria for screening.

RESULTS: The experimental results show that the inpainting effect of the proposed method is significantly improved in subjective vision, and the structural similarity (SSIM), peak signal-to-noise ratio (PSNR), and inpainting speed of the inpainting results are all improved.

CONCLUSION: For different types of images, the proposed method has a better inpainting effect and higher inpainting speed than the other three advanced methods.

Transformer-Guided Video Inpainting Algorithm Based on Local Spatial-Temporal joint

Jing Wang, ZongJu Yang — Tue, 15 Aug 2023 00:00:00 +0000

INTRODUCTION: Video inpainting is a very important task in computer vision, and it’s a key component of various practical applications. It also plays an important role in video occlusion removal, traffic monitoring and old movie restoration technology. Video inpainting is to obtain reasonable content from the video sequence to fill the missing region, and maintain time continuity and spatial consistency.
OBJECTIVES: In previous studies, due to the complexity of the scene of video inpainting, there are often cases of fast motion of objects in the video or motion of background objects, which will lead to optical flow failure. So the current video inpainting algorithm hasn’t met the requirements of practical applications. In order to avoid the problem of optical flow failure, this paper proposes a transformer-guided video inpainting model based on local Spatial-temporal joint.
METHODS: First, considering the rich Spatial-temporal relationship between local flows, a Local Spatial-Temporal Joint Network (LSTN) including encoder, decoder and transformer module is designed to roughly inpaint the local corrupted frames, and the Deep Flow Network is used to calculate the local bidirectional corrupted flows. Then, the local corrupted optical flow map is input into the Local Flow Completion Network (LFCN) with pseudo 3D convolution and attention mechanism to obtain a complete set of bidirectional local optical flow maps. Finally, the roughly inpainted local frame and the complete bidirectional local optical flow map are sent to the Spatial-temporal transformer and the inpainted video frame is output.
RESULTS: Experiments show that the algorithm achieves high quality results in the video target removal task, and has a certain improvement in indicators compared with advanced technologies.
CONCLUSION: Transformer-Guided Video Inpainting Algorithm Based on Local Spatial-Temporal joint can obtain high-quality optical flow information and inpainted result video.

EEG Emotion Recognition based on Multi scale Self Attention Convolutional Networks

Hao Chao, Fang Yuan — Wed, 06 Sep 2023 00:00:00 +0000

A multi-view self-attention module is proposed and paired with a multi-scale convolutional model to build
a multi-view self-attention convolutional network for multi-channel EEG emotion recognition. First, time
and frequency domain characteristics are extracted from multi-channel EEG signals, and a three-dimensional
feature matrix is built using spatial mapping connections. Then, a multi-scale convolutional network extracts
the high-level abstract features from the feature matrix, and a multi-view self-attention network strengthens
the features. Finally, use the multilayer perceptron for sentiment classification. The experimental results reveal
that the multi-view self-attention convolutional network can effectively integrate the time domain, frequency
domain, and spatial domain elements of EEG signals using the DEAP public emotion dataset. The multi-view
self-attention module can eliminate superfluous data, apply attention weight to the network to hasten network
convergence, and enhance model recognition precision.

A Comparison of SES and SMA Method Against Production Level Property of Fabrication Precision Engineering and Its Effect on Production Planning (Case Study PT X)

Via Rensi Novita Alfa Reza, Ancala Laras Putri — Wed, 27 Sep 2023 00:00:00 +0000

One of the goals of forecasting is to be able to predict the data needed in the future, one of which is the data on demand for the amount of production in a company. PT X is a manufacturing company based on demand or custom. This leads to uncertainty in the use of required materials, so proper forecasting is needed to estimate the material stock requirements. This study used single exponential smoothing and single moving average methods with quantitative approaches to aluminum materials. By calculating forecasting using these two methods, it is possible to find the best method for use by PT X. Based on the test, the SES method has the smallest error rate so it can be used to analyze the data, with α=0.8 yielding a forecast in the 13th month of 308.71408 pcs.

The Power of AI-Assisted Diagnosis

Jiaji Wang — Wed, 06 Sep 2023 00:00:00 +0000

The rapid advancements in artificial intelligence (AI) have unleashed a wave of transformative technologies, and one area that has witnessed significant progress is AI-assisted diagnosis in healthcare. With the ability to analyze vast amounts of medical data, learn from patterns, and make accurate predictions, AI systems hold immense potential to revolutionize the diagnostic process, enabling earlier detection, improved accuracy, and personalized treatment recommendations. This review aims to explore the impact of AI in healthcare, specifically focusing on its role in assisting physicians with diagnosis, highlighting the benefits, challenges, and ethical considerations associated with the integration of AI systems into clinical practice. Through the utilization of AI's capabilities, the enhancement of patient outcomes, optimization of resource allocation, and the reshaping of medical professionals' approaches to diagnosis and treatment can be achieved.