| Summary: | 碩士 === 國立高雄應用科技大學 === 資訊工程系 === 104 === Huge benefits can be obtained by mining information from Big Data. Analyzing large volumes of consumption behavior data that are limited by conventional machine learning techniques and computational analysis becomes a critical problem as Big Data is examined. Furthermore, there is a need for powerful visual-based analytics tools when pictures have become a core content component on the Internet. Hence, in this study, we explore Deep Learning with convolutional neural networks with a goal of resolving clothing style classification and retrieval tasks. To reduce training complexity, transfer learning is incorporated by fine-tuning pre-trained models on large scale datasets. Furthermore, because the parameters are vast for any given deep net, one architecture inspired from Adaboost is designed to use multiple deep nets that are trained with a sub-dataset. Thus, the training time can be accelerated if each net is computed in one client node in a distributed computing environment. Moreover, to increase system flexibility, two architectures with multiple deep nets with two outputs are proposed for binary-class classification. Therefore, when new classes are added, no additional computation is needed for all training data. Experiments are performed to compare existing systems with hand-crafted features and conventional learning models. According to the results, the proposed system can provide significant improvements on three public clothing datasets for style classifications.
|
|---|
