Trained segmentation, classification, and style transformation models are used to apply style effects to image segments corresponding to objects (e.g., a portrait of a person) present in an image. When depth information is included with a source image it may be used to segment, classify, and/or apply style transformation to an image or image segments (as the case may be). One or more of the segmentation, classification, and style transformations may be performed at a computing apparatus or as a service for example, in the cloud. A model management tool implemented at a server(s) may continuously train models using supervised or unsupervised training techniques to improve the quality of segmentation, classification and style transformation as well as add new capabilities. Updated models may be pushed to services offering image processing. Additionally or alternatively, a mobile application can inquire about versions of models and request updated models, or in another case, the model management tool may push updates to the mobile application. Thus, the mobile application may perform on-board image processing including segmentation, classification, and style transformation.
展开▼