Machine learning applications in detecting rip channels from images

Maryan, Corey; Michael, Christopher; Ioup, Elias; Abdelguerfi, Mahdi

doi:10.1016/j.asoc.2019.02.017

Cited by 21 publications

(16 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This feature is one of the methods of Viola-Jones [27]. Haar-like features are rectangular features [28], which can be given a specific indication of an image or image [29]. Features like Haar are used to identify objects based on the simple value of a feature, not pixel values contained from the image of the object [30] [31].…”

Section: B Cascade Classifiermentioning

confidence: 99%

Face Shape-Based Physiognomy in LinkedIn Profiles with Cascade Classifier and K-Means Clustering

Purwono

Ma’arif

Wulandari

2021

2021 8th International Conference on Electrical Engineering, Computer Science and Informatics (EECSI)

View full text Add to dashboard Cite

The progress of a company is influenced by the excellent performance of its employee. The recruitment process should be done in a correct procedure so that it would not have the potential to harm the company. The improved use of social media can be an aspect to be applied in a recruitment process. LinkedIn is a social media platform that has many users which focuses on the career development aspect. Profile photos are commonly used in social media. In physiognomy, a personality analysis can be carried out based on his/her outward appearance. The profile photo can be an aspect of personality analysis with this knowledge. This research aimed to predict the face shape based on LinkedIn profile photos. A Cascade classifier algorithm with a haar-like feature was used to detect the face area. Dlib library was used to detect face landmarks. K-Means algorithm was used to differentiate the border of hair and facial skin. Indicators of the face shape calculation are the value of face angle, which is the arctangent of the face landmarks matrix; similarity value from the standard deviation calculation between horizontal line 1, 2, and 3; and diameter value which resulted from the standard deviation calculation between horizontal line 2 and vertical line 4. We provide output as face shape from the LinkedIn profile photos. Based on ten profile photo samples, only two predictions were incorrect.

show abstract

Section: B Cascade Classifiermentioning

confidence: 99%

Face Shape-Based Physiognomy in LinkedIn Profiles with Cascade Classifier and K-Means Clustering

Purwono

Ma’arif

Wulandari

2021

2021 8th International Conference on Electrical Engineering, Computer Science and Informatics (EECSI)

View full text Add to dashboard Cite

show abstract

“…Then, the augmented feature vector was used for training the final level of the learner or the metaclassifier [30][31] [32].…”

Section: Models and Parametersmentioning

confidence: 99%

Flight Data of Airplane for Wind Forecasting

Sharma

Hoque

Ioup

et al. 2020

IGARSS 2020 - 2020 IEEE International Geoscience and Remote Sensing Symposium

Self Cite

View full text Add to dashboard Cite

Department of Computer Science for constantly guiding me and encouraging me to work hard for the past two years. I am very grateful to Dr. Mahdi for the support and motivation he has given me through every stage of my research. I am thankful to Dr. Hoque for providing me an opportunity to learn and grow my knowledge in different areas of Machine Learning and helping me find the answers to my research questions.

show abstract

“…for non-stationary rip currents. While Timex images are most commonly viewed by human experts, Maryan et al [35] trained a machine learning model to determine whether a Timex image contains a rip channel or not. They reported a detection rate of 85% for various beach locations.…”

mentioning

confidence: 99%

“…The human experts indicated that the visualizations were valuable and contained Currently, there are two rip detection methods that utilize machine learning methods. Maryan et al [35] employ a Viola-Jones framework [54] to train their model to detect rips from Timex images (see Timex column in Table 1), while de Silva et al [37] used a modified deep learning technique…”

mentioning

confidence: 99%

See 1 more Smart Citation

Flow-Based Rip Current Detection and Visualization

et al. 2022

View full text Add to dashboard Cite

Rip currents are like rivers of fast-moving water that can quickly carry the unwary out to sea. They are not easy to recognize, especially to the untrained observer. Other than in-situ current measurements, there exists a number of methods that analyzes images and videos to detect rip currents. Most of these techniques base their detection on the appearance of rip currents such as foamy pattern, discoloration of water, and locations of breaking waves. Leading methods use either image processing or machine learning of images and/or video input. In this paper, we analyze the behavior of water movement rather than simply its appearance to detect rip currents. Specifically, we investigated several flow visualization methods and tune them to detect rip currents. Based on our study, we recommend two methods that allowed us to detect rip currents where other methods have failed. And because the methods originated as visualization techniques, any presence of rip currents are automatically highlighted. We also evaluated these two methods against previously annotated results by rip current experts, and found that our detections were sufficiently sensitive that some expert annotations were relabeled.

show abstract

Machine learning applications in detecting rip channels from images

Cited by 21 publications

References 8 publications

Face Shape-Based Physiognomy in LinkedIn Profiles with Cascade Classifier and K-Means Clustering

Face Shape-Based Physiognomy in LinkedIn Profiles with Cascade Classifier and K-Means Clustering

Flight Data of Airplane for Wind Forecasting

Flow-Based Rip Current Detection and Visualization

Contact Info

Product

Resources

About