Weather Classification
Anish Sudhir Ghiya, Vaibhav Vijay, Aditi Ranganath, Prateek Chaturvedi, Sharmila Banu Kather, Bala Krushna Tripathy · 2023
Weather conditions are essential to know what is about to happen. They often disrupt day-to-day activities. In the field of computer vision, it is a difficult task due to the high diversity and the lack of distinguishable features that can be extracted from them. In this chapter, we propose an Image Embedding based Stacked Generalization (IESG) method to determine the best image embeddings that can be generated from a given image. A convolutional autoencoder is an unsupervised technique for feature extraction. Multiple machine learning models are trained to generate a stacked ensemble model using StackNet architecture. Also, the method proposed in the chapter improves predictions by 5% when compared to the original paper which used this data set.