ICACT20220350 Question.1
Questioner: phamdinhlam@kgu.ac.kr 2022-02-13 오후 2:01:09 |
ICACT20220350 Answer.1
Answer by Auhor jypark9740@g.skku.edu 2022-02-13 오후 2:01:09
Chrome  Click!! |
| Could you explain more about the data set used in the paper for evaluation? How many features, labels of fault product? And how did you apply the Xception model with the data set? |
In this paper, an experiment was conducted using image data of a submersible pump impeller among casting products. The open data set was used for the experiment, and the train data were 3,758 defective images and 2,875 normal image data. And for test data, 453 defective images and 262 normal image data were used.
There are two types of casting data: normal and defective products. Although various defects exist in the abnormal casting data, the label of the defect and the defect segmentation are not performed. In this experiment, data preprocessing was performed by dividing it into normal data and abnormal data.
Using the above data, Xception was used as a deep learning model for product classification. The model structure is composed of three flows: Entry flow, Middle flow, and Exit flow. In entry flow, there are 2 Convolution Layers and 6 Separable Conv Layers.
Middle flow consists of 3 Separable Layers and repeats 8 times.
In Exit flow, 4 Separable Layers are configured. Behind all convolutional and separable layers is batch normalization. And the model was constructed using Maxpooling and ReLu functions. |
ICACT20220350 Question.2
Questioner: tomayoon@ieee.org 2022-02-14 오전 10:26:53 |
ICACT20220350 Answer.2
Answer by Auhor jypark9740@g.skku.edu 2022-02-14 오전 10:26:53
Chrome Click!! |
| First of all, thank you for the good research paper and presentation contents. I would like to know more about the research environment used in this study, and also in detail about the equipment & operating systems, development languages and software libraries you used. |
The research environments used in this study are as follows. Hardware consists of an Intel Core i7-8700K processor, a computer with GTX 1080 TI and 12GB RAM. And the software environment has experimented with Windows Tensorflow 2.0 Framework and Python 3.7 environment. |
ICACT20220350 Question.3
Questioner: faryal.nosheen@ceme.nust.edu.pk 2022-02-14 오후 4:08:24 |
ICACT20220350 Answer.3
Answer by Auhor jypark9740@g.skku.edu 2022-02-14 오후 4:08:24
Chrome Click!! |
| You have chosen Gaussian noise filter for the datasets. Could you please explain how you decide to apply this particular filter. |
Thank you for your good question. The Gauss filter is a linear filter. Usually used to blur images or reduce noise. If use two things to blur and remove, you can use it for edge detection. The kind of noise filter of the image is very much and various. Future studies will apply a variety of noise filters and to improve image. |
ICACT20220350 Question.4
Questioner: kakikura_sayoko@sec.ee.kagu.tus.ac.jp 2022-02-14 오후 1:29:50 |
ICACT20220350 Answer.4
Answer by Auhor jypark9740@g.skku.edu 2022-02-14 오후 1:29:50
Chrome Click!! |
| What kind of images were not classified correctly?
Are there any common features in the images that were classified as FP/FN? |
When learning, the specific casting data defect portion to which noise is applied is that the image is not accurately classified due to noise. The defect of noise is not confirmed by the naked eye. Future studies will conduct noise filter-specific comparison experiments and will be experimented with the impact on accuracy in some cases. |
ICACT20220350 Question.5
Questioner: shenhanji@cnic.cn 2022-02-15 오전 10:57:25 |
ICACT20220350 Answer.5
Answer by Auhor jypark9740@g.skku.edu 2022-02-15 오전 10:57:25
Chrome Click!! |
| Thanks for your presentation！What is your experimental data set? What problems does your research solve in the real environment? |
In this paper, an experiment was conducted using image data of a submersible pump impeller among casting products. When evaluating the quality of the product in the actual manufacturing site, the quality can be evaluated by improving data with noise and increasing accuracy. |
ICACT20220350 Question.6
Questioner: namacabale@gmail.com 2022-02-15 오후 1:42:55 |
ICACT20220350 Answer.6
Answer by Auhor jypark9740@g.skku.edu 2022-02-15 오후 1:42:55
Chrome Click!! |
| Dear Author, you have a very good research. Is there an existing technology almost similar to your proposal that has been deployed in real world environment? What does it lack that your work specifically addressed? Thank you and stay safe.
|
The method of improve denoising and resolution in the real world environment is diverse, and its methods are applied. In this paper, we proposed technologies using wavelet denoising, but we will conduct a variety of methods of research in future studies. |
ICACT20220350 Question.7
Questioner: namacabale@gmail.com 2022-02-16 오후 12:58:28 |
ICACT20220350 Answer.7
Answer by Auhor jypark9740@g.skku.edu 2022-02-16 오후 12:58:28
Chrome Click!! |
| Dear Author, you have a very good research. Is there an existing technology almost similar to your proposal that has been deployed in real world environment? What does it lack that your work specifically addressed? Thank you and stay safe. |
Thank you for your good question. The actual site knows that more existing techniques are being applied. What is specifically addressed in my paper is a denoising filter and a lightweight Xception model structure. |
