A hybrid deep learning skin cancer prediction framework

Engineering Science and Technology an International Journal, Год журнала: 2024, Номер 57, С. 101818 - 101818

Опубликована: Авг. 27, 2024

Язык: Английский

---

AdvancingTire Safety: Explainable Artificial Intelligence-Powered Foreign Object Defect Detection with Xception Networks and Grad-CAM Interpretation

Applied Sciences, Год журнала: 2024, Номер 14(10), С. 4267 - 4267

Опубликована: Май 17, 2024

Automatic detection of tire defects has become an important issue for production companies since these cause road accidents and loss human lives. Defects in the inner structure cannot be detected with naked eye; thus, a radiographic image is gathered using X-ray cameras. This then examined by quality control operator, decision made on whether it defective or not. Among all defect types, foreign object type most common may occur anywhere tire. study proposes explainable deep learning model based Xception Grad-CAM approaches. was fine-tuned trained novel real dataset consisting 2303 tires 49,198 non-defective. The class augmented custom augmentation technique to solve imbalance problem dataset. Experimental results show that proposed detects objects accuracy 99.19%, recall 98.75%, precision 99.34%, f-score 99.05%. provided clear advantage over similar literature studies.

Язык: Английский

---

Attention-based deep learning for tire defect detection: Fusing local and global features in an industrial case study

Expert Systems with Applications, Год журнала: 2025, Номер unknown, С. 126473 - 126473

Опубликована: Янв. 1, 2025

Язык: Английский

---

Ball Screw Drive Surface Defect Model Based on Transfer Learning Approach

Lecture notes in networks and systems, Год журнала: 2025, Номер unknown, С. 451 - 457

Опубликована: Янв. 1, 2025

Язык: Английский

---

Hybrid twin attention based convolutional stacked sparse autoencoder for classification of defected weld images

Computers & Electrical Engineering, Год журнала: 2025, Номер 124, С. 110328 - 110328

Опубликована: Апрель 17, 2025

Язык: Английский

---

Automated explainable deep learning framework for multiclass skin cancer detection and classification using hybrid YOLOv8 and vision transformer (ViT)

Biomedical Signal Processing and Control, Год журнала: 2025, Номер 108, С. 107934 - 107934

Опубликована: Апрель 29, 2025

Язык: Английский

---

Research on tire appearance defect detection algorithm based on efficient multi-scale convolution

Measurement Science and Technology, Год журнала: 2024, Номер 36(1), С. 015009 - 015009

Опубликована: Окт. 8, 2024

Abstract Due to the large randomness of tire appearance defect size and complex diverse shapes, existing target detection algorithm is prone missing misidentifying targets, accuracy limited, model large, which not conducive deployment on embedded devices. In this paper, efficient multi-scale convolution (EMC) mode proposed, C2f-EMC module designed basis, improves network structure YOLOv8, detection, reduces number parameters in model. EMC first divides input feature images into four parts average carries out with cores 1 × 1, 3 3, 5 7 sizes respectively. Then, obtained results are stacked, cross-channel fusion realized by point-by-point convolution. After determining C2f-EMC, best improvement position determined through comparative experiments. Experiments show that after above improvements, parameter reduced 4.85%, calculation amount 2.82%, 4.44%, recall rate 2.8%, mAP50 1.0%, mAP50-95 1.3%, F1 2%. The task can be completed more accurately requirements devices better met.

Язык: Английский

---

Comprehensive analysis study of techniques in different domains for Turkish music genre classification task

Neural Computing and Applications, Год журнала: 2024, Номер 37(5), С. 3005 - 3021

Опубликована: Дек. 10, 2024

Язык: Английский

---