Uniform Experimental Design for Optimizing the Parameters of Multi-input Convolutional Neural Networks

Lin, Cheng‐Jian; Wu, Chao‐Hsiung; Sun, Chi‐Chia; Lin, Chu‐Hsing

doi:10.18494/sam.2020.2771

Cited by 3 publications

(2 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To further enhance the efficiency of the manufacturing process, we propose to perform image reprocessing and utilize convolutional neural networks (CNNs) (3)(4)(5)(6)(7) in this study, we achieve an accuracy above 95%, maintain a model file size of less than 20 MB, and process each image for a prediction time of less than 30 ms.…”

Section: Introductionmentioning

confidence: 99%

Clamp Dot Image Classification Using Neural Network

Srikam,

Saenthon

2024

Sensors and Materials

View full text Add to dashboard Cite

In this paper, we discuss the classification of images captured by a machine camera while assembling components. To crop out specific points of interest, we employ image processing. Additionally, we utilize deep learning techniques, specifically convolutional neural networks, to identify the type of equipment being assembled. This approach allows us to determine and record specific parts within a device. However, the main challenge of this project is to achieve both high accuracy and the shortest possible prediction time.

show abstract

Section: Introductionmentioning

confidence: 99%

Clamp Dot Image Classification Using Neural Network

Srikam,

Saenthon

2024

Sensors and Materials

View full text Add to dashboard Cite

show abstract

“…Usually, the above fingerprints do not contain enough details, and it is difficult for the conventional matching methods based on the details to achieve ideal results. Although most electronic consumer terminals cannot provide enough computing power for AI-based image matching algorithms, (1)(2)(3)(4)(5)(6) image matching based on local feature points is widely used in the above scenes, such as the well-known scale-invariant feature transform (SIFT) algorithm, (7) the speeded-up robust features (SURF) algorithm, (8) binary robust independent elementary features (BRIEF) algorithm, (9) and the features from accelerated segment test (FAST) and rotated BRIEF (ORB) algorithm. (10) However, differences between the various image matching algorithms based on feature points, such as orientation, coordinates, and sub-vector description, may lead to diverse effects on the same image.…”

Section: Introductionmentioning

confidence: 99%

Image Matching Algorithm Based on Topology Consistency of Bidirectional Optimal Matching Point Pairs

Wu¹,

Chen²,

Bian³

et al. 2022

Sensors and Materials

View full text Add to dashboard Cite

The random sample consensus (RANSAC) algorithm is commonly used to estimate the parameters of the image transformation model based on matching point pairs in the featurebased image matching field. If the dataset of matching point pairs contains outliers, the conventional RANSAC algorithm may take a large number of iterations to obtain the desired model. To reduce mismatching, we propose the bidirectional optimal matching method, aiming to find robust parameters within a short time. The topology-consistency-based sampling method is introduced to divide the dataset into certain consensus sets, and sampling from each of them can reduce randomness. Then, all point pairs from a consensus set are used to estimate a model, and a point pair unsuitable for the model is deleted in each iteration, which is demonstrated to be faster than the conventional RANSAC. The superiority of the proposed method in fingerprint matching based on the scale-invariant feature transform is shown in experiments.

show abstract

Customized Modification of Welan Gum Properties Through Controllable Grafting of Acrylamide

Xu,

Li,

et al. 2024

Biopolymers

View full text Add to dashboard Cite

Welan gum (WG) has a wide range of applications, but it is not yet suitable for applications such as oil recovery profile control that have complex requirements for viscosity, gelation properties, and so forth. Grafting modification is an important strategy for improving the property of WG, but there are few reports on controllable modification of WG to customize it for specific application. Acrylamide (AM) dosage was identified as the key factor affecting the grafting ratio of AM onto WG by a uniform experimental design. The grafting ratio can be directly adjusted between 99% and 378% based on the positive correlation with dosage of AM, and viscosity can be adjusted between 206 and 327 mPa s based on the negative correlation with grafting ratio. The 50% weight loss temperature of W11 with a grafting ratio of 110% raised from 314 to 336°C after grafting. The viscosity of the hydrogel formed with WG11 reached 15,654 mPa s, nearly nine times higher than that of unmodified WG. In addition, the gelation time can be controlled within 5 days, so that it can be injected to the optimal area in oilfield profile, avoiding pipeline blockage. This study enables adjusting viscosity of WG grafted with AM by controlling the grafting rate, and enhances gelation performance and thermal stability of WG, which will expand the application of WG in oil recovery and other fields.

show abstract

Uniform Experimental Design for Optimizing the Parameters of Multi-input Convolutional Neural Networks

Cited by 3 publications

References 17 publications

Clamp Dot Image Classification Using Neural Network

Clamp Dot Image Classification Using Neural Network

Image Matching Algorithm Based on Topology Consistency of Bidirectional Optimal Matching Point Pairs

Customized Modification of Welan Gum Properties Through Controllable Grafting of Acrylamide

Contact Info

Product

Resources

About