Deep salient object segmentation

Abstract

Systems, methods, and non-transitory computer-readable media are disclosed for segmenting objects in digital visual media utilizing one or more salient content neural networks. In particular, in one or more embodiments, the disclosed systems and methods train one or more salient content neural networks to efficiently identify foreground pixels in digital visual media. Moreover, in one or more embodiments, the disclosed systems and methods provide a trained salient content neural network to a mobile device, allowing the mobile device to directly select salient objects in digital visual media utilizing a trained neural network. Furthermore, in one or more embodiments, the disclosed systems and methods train and provide multiple salient content neural networks, such that mobile devices can identify objects in real-time digital visual media feeds (utilizing a first salient content neural network) and identify objects in static digital images (utilizing a second salient content neural network).

Description

Background technique [0001] In recent years, the use of digital visual media on client computing devices has seen a dramatic increase. Indeed, individuals and businesses are increasingly using laptops, tablets, smartphones, handheld devices, and other mobile technologies for a variety of tasks involving digital visual media. For example, individuals and businesses are increasingly utilizing smartphones to capture, view and modify digital visual media, such as portrait images, "selfies" or digital videos. [0002] Although conventional digital visual media systems allow users to capture and modify digital visual media, they also have a number of significant disadvantages. For example, conventional digital visual media systems can utilize cameras to capture digital visual media, but cannot easily, quickly, or efficiently select or separate individual objects from other pixels depicted in the digital visual media. [0003] Some traditional digital visual media systems assist us...

AI Technical Summary

Problems solved by technology

Furthermore, traditional digital visual media systems utilizing classification models cannot operate in real time across multiple digital images

For example, traditional digital visual media systems cannot segment objects depicted in live digital visual media feeds (e.g., live video feeds from smartphone cameras)

Images