Efficient Deep Learning Methods for Identification of Defective Casting Products

Bolla, Bharath Kumar; Kingam, Mohan; Ethiraj, Sabeesh

doi:10.1007/978-3-031-22405-8_13

Cited by 5 publications

(3 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Data augmentation is a widely used technique for improving the performance of machine learning models [9], [10], [11]. However, there are cases where augmentation can decrease model performance, particularly in detecting defects [12], [13]. To address this, recent developments in deep learning have led to the use of neural network models for generating synthetic data.…”

Section: Synthetic Data For Images and Textmentioning

confidence: 99%

Evaluating the Utility of GAN Generated Synthetic Tabular Data for Class Balancing and Low Resource Settings

Chereddy,

Bolla

2023

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

The present study aimed to address the issue of imbalanced data in classification tasks and evaluated the suitability of SMOTE, ADASYN, and GAN techniques in generating synthetic data to address the class imbalance and improve the performance of classification models in low-resource settings. The study employed the Generalised Linear Model (GLM) algorithm for class balancing experiments and the Random Forest (RF) algorithm for low-resource setting experiments to assess model performance under varying training data. The recall metric was the primary evaluation metric for all classification models. The results of the class balancing experiments showed that the GLM model trained on GAN-balanced data achieved the highest recall value. Similarly, in low-resource experiments, models trained on data enhanced with GAN-synthesized data exhibited better recall values than original data. These findings demonstrate the potential of GAN-generated synthetic data for addressing the challenge of imbalanced data in classification tasks and improving model performance in low-resource settings.

show abstract

Section: Synthetic Data For Images and Textmentioning

confidence: 99%

Evaluating the Utility of GAN Generated Synthetic Tabular Data for Class Balancing and Low Resource Settings

Chereddy,

Bolla

2023

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…In [13], an enhanced domain adaptive Faster R-CNN model was introduced with its superior capability to detect void and inclusion defects in spacecraft composite structures (SCSs). In [14], different pre-trained and custom-built architectures were compared and contrasted with model size, performance In [13], an enhanced domain adaptive Faster R-CNN model was introduced with its superior capability to detect void and inclusion defects in spacecraft composite structures (SCSs). In [14], different pre-trained and custom-built architectures were compared and contrasted with model size, performance, and CPU latency in detecting defective casting products.…”

Section: Introductionmentioning

confidence: 99%

“…In [14], different pre-trained and custom-built architectures were compared and contrasted with model size, performance In [13], an enhanced domain adaptive Faster R-CNN model was introduced with its superior capability to detect void and inclusion defects in spacecraft composite structures (SCSs). In [14], different pre-trained and custom-built architectures were compared and contrasted with model size, performance, and CPU latency in detecting defective casting products. In [15], the problem of identifying small defects during an industrial inspection was defined.…”

Section: Introductionmentioning

confidence: 99%

Transfer Learning for Detection of Casting Defects Model In Scope of Industrial 4.0

TANYILDIZ,

BATUR ŞAHİN

2023

Türk Doğa Ve Fen Dergisi

View full text Add to dashboard Cite

Casting represents a production process where a liquid material is poured into a mold with a hollow cavity, usually of the intended shape, following which its solidification is allowed. Numerous defect types are available, including blow holes, pin holes, burrs, mold material defects, shrinkage defects, metallurgical defects, casting metal defects, etc. All industries have quality control departments to eliminate the occurrence of this defective product. But the main problem is that this inspection process is done manually. This is a very time consuming process and due to human sensitivity this is not 100% accurate. In this study, we will verify whether the "manual inspection" bottleneck can be eliminated by automating the inspection process with transfer learning in the manufacturing process of casting products. In this study, we will verify whether the "manual inspection" bottleneck can be eliminated by automating the inspection process with transfer learning in the manufacturing process of casting products. In this study, the casting images were divided into two separate classes, and the classification process was carried out by applying deep learning architectures. The benefits of this proposed approach are discussed and proposed as a more efficient way to control the quality of final products under Industry 4.0.

show abstract

Smoke recognition in steelmaking converter images: Class-distance-based feature selection model

Wu,

Long,

Wang

et al. 2024

Expert Systems with Applications

View full text Add to dashboard Cite

Efficient Deep Learning Methods for Identification of Defective Casting Products

Cited by 5 publications

References 13 publications

Evaluating the Utility of GAN Generated Synthetic Tabular Data for Class Balancing and Low Resource Settings

Evaluating the Utility of GAN Generated Synthetic Tabular Data for Class Balancing and Low Resource Settings

Transfer Learning for Detection of Casting Defects Model In Scope of Industrial 4.0

Smoke recognition in steelmaking converter images: Class-distance-based feature selection model

Contact Info

Product

Resources

About