2021 IEEE International Conference on Power Electronics, Computer Applications (ICPECA) 2021
DOI: 10.1109/icpeca51329.2021.9362627
|View full text |Cite
|
Sign up to set email alerts
|

An improvement and application of a model conducive to productivity optimization

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
2
0

Year Published

2023
2023
2024
2024

Publication Types

Select...
3

Relationship

0
3

Authors

Journals

citations
Cited by 3 publications
(7 citation statements)
references
References 7 publications
0
2
0
Order By: Relevance
“…In industries, datasets are generally proprietary and are not exposed to researchers, due to the costs of generating them, for example, BS5-DET [6], but it is still possible to find datasets already created and for open use, like the DAGM dataset [25] or the COCO dataset [26]. Most free datasets are found in Kaggle [27].…”
Section: Resultsmentioning
confidence: 99%
See 2 more Smart Citations
“…In industries, datasets are generally proprietary and are not exposed to researchers, due to the costs of generating them, for example, BS5-DET [6], but it is still possible to find datasets already created and for open use, like the DAGM dataset [25] or the COCO dataset [26]. Most free datasets are found in Kaggle [27].…”
Section: Resultsmentioning
confidence: 99%
“…Metal surfaces are the group with the most subtypes because metals can be used in their pure state, in alloys such as steel (the most used metal in this systematic review), or in interesting surfaces such as titanium-coated metal [31], microscopic images from thin metal film in electronic components [67], semiconductor wafers (from the metal layers) [71], polishing metal shafts [29], car wiper arms [58], microscopic metal parts [62], cuts from laser cutting machines [64], wind turbine blades [66], insulators in the transmission line aims [68], and X-ray images from metals [72]. Most of the datasets are created by the authors and are kept private; however, some are free to use like BS5-DET [6], CSU_STEEL [41], and GC10-DET [46]. For studies that aim to compare the proposed methods with traditional networks or their datasets such as [41,46,50,53,57,84], the most commonly used option for metal is NEU-DET [28], which has been tested and contains six of the most common defects (crazing, inclusion, patches, pitted surface, rolled-in scale, and scratches).…”
Section: Metalmentioning
confidence: 99%
See 1 more Smart Citation
“…Non-destructive techniques and photogrammetry have been used for the inspection of brick buildings, but manual inspection is not time-effective, and it can lead to errors. The presented detection methods reduce the time needed for the work to be completed [10]. Transfer learning was used to address this gap by classifying brickwork images as cracked or normal.…”
Section: Related Workmentioning
confidence: 99%
“…Currently, manual and visual inspections performed by experts are expensive due to the high cost of human labor during working hours, the possibility of material waste, and the degraded quality of shipped products [ 1 ]. In contrast, the use of machine learning algorithms for automatic defect detection reduces labor consumption [ 2 ]. In recent years, automatic defect detection has played a critical role in the industry’s inspection process.…”
Section: Introductionmentioning
confidence: 99%