Classifying the visual features in images to retrieve a specific image is a significant problem within the computer vision field especially when dealing with historical faded colored images.Thus,there were lots of eff...Classifying the visual features in images to retrieve a specific image is a significant problem within the computer vision field especially when dealing with historical faded colored images.Thus,there were lots of efforts trying to automate the classification operation and retrieve similar images accurately.To reach this goal,we developed a VGG19 deep convolutional neural network to extract the visual features from the images automatically.Then,the distances among the extracted features vectors are measured and a similarity score is generated using a Siamese deep neural network.The Siamese model built and trained at first from scratch but,it didn’t generated high evaluation metrices.Thus,we re-built it from VGG19 pre-trained deep learning model to generate higher evaluation metrices.Afterward,three different distance metrics combined with the Sigmoid activation function are experimented looking for the most accurate method formeasuring the similarities among the retrieved images.Reaching that the highest evaluation parameters generated using the Cosine distance metric.Moreover,the Graphics Processing Unit(GPU)utilized to run the code instead of running it on the Central Processing Unit(CPU).This step optimized the execution further since it expedited both the training and the retrieval time efficiently.After extensive experimentation,we reached satisfactory solution recording 0.98 and 0.99 F-score for the classification and for the retrieval,respectively.展开更多
The characteristics and climatology of funnel clouds in Alaska were examined using operational radiosondes, surface meteorological observations, and reanalysis data. Funnel clouds occurred under weak synoptic forcing ...The characteristics and climatology of funnel clouds in Alaska were examined using operational radiosondes, surface meteorological observations, and reanalysis data. Funnel clouds occurred under weak synoptic forcing between May and September between 11 am and 6 pm Alaska Daylight Time with a maximum occurrence in July. They occurred under Convective Available Potential Energy >500 J·kg-1 and strong low-level wind shear. Characteristic atmospheric profiles during funnel cloud events served to develop a retrieval algorithm based on similarity testing. Out of more than 129,000 soundings between 1971 and 2014, 2724, 442, and 744 profiles were similar to the profiles of observed funnel cloud events in the Interior, Alaska West Coast, and Anchorage regions. While the number of reported funnel clouds has increased since 2000, the frequency of synoptic situations favorable for such events has decreased.展开更多
基金The authors would like to thank the Deanship of Scientific Research at Umm Al-Qura University for supporting this work by Grant Code:(22UQU4400271DSR01).
文摘Classifying the visual features in images to retrieve a specific image is a significant problem within the computer vision field especially when dealing with historical faded colored images.Thus,there were lots of efforts trying to automate the classification operation and retrieve similar images accurately.To reach this goal,we developed a VGG19 deep convolutional neural network to extract the visual features from the images automatically.Then,the distances among the extracted features vectors are measured and a similarity score is generated using a Siamese deep neural network.The Siamese model built and trained at first from scratch but,it didn’t generated high evaluation metrices.Thus,we re-built it from VGG19 pre-trained deep learning model to generate higher evaluation metrices.Afterward,three different distance metrics combined with the Sigmoid activation function are experimented looking for the most accurate method formeasuring the similarities among the retrieved images.Reaching that the highest evaluation parameters generated using the Cosine distance metric.Moreover,the Graphics Processing Unit(GPU)utilized to run the code instead of running it on the Central Processing Unit(CPU).This step optimized the execution further since it expedited both the training and the retrieval time efficiently.After extensive experimentation,we reached satisfactory solution recording 0.98 and 0.99 F-score for the classification and for the retrieval,respectively.
基金the National Science Foundation(NSF),the SOARS program,the Gwichyaa Zhee Gwich’in Tribal Government,and SLOAN for financial support.
文摘The characteristics and climatology of funnel clouds in Alaska were examined using operational radiosondes, surface meteorological observations, and reanalysis data. Funnel clouds occurred under weak synoptic forcing between May and September between 11 am and 6 pm Alaska Daylight Time with a maximum occurrence in July. They occurred under Convective Available Potential Energy >500 J·kg-1 and strong low-level wind shear. Characteristic atmospheric profiles during funnel cloud events served to develop a retrieval algorithm based on similarity testing. Out of more than 129,000 soundings between 1971 and 2014, 2724, 442, and 744 profiles were similar to the profiles of observed funnel cloud events in the Interior, Alaska West Coast, and Anchorage regions. While the number of reported funnel clouds has increased since 2000, the frequency of synoptic situations favorable for such events has decreased.
