Context-Aware Synthesis and Placement of Object Instances

TL;DR

This paper introduces an end-to-end neural network that learns to insert object instances into images coherently by jointly determining their location, scale, shape, and appearance, enhancing image editing capabilities.

Contribution

It presents a novel neural network architecture with two generative modules for object placement and appearance, trained jointly with supervised and unsupervised data.

Findings

Effective insertion of objects at diverse locations

Ability to generate varied object appearances

Outperforms strong baseline methods

Abstract

Learning to insert an object instance into an image in a semantically coherent manner is a challenging and interesting problem. Solving it requires (a) determining a location to place an object in the scene and (b) determining its appearance at the location. Such an object insertion model can potentially facilitate numerous image editing and scene parsing applications. In this paper, we propose an end-to-end trainable neural network for the task of inserting an object instance mask of a specified class into the semantic label map of an image. Our network consists of two generative modules where one determines where the inserted object mask should be (i.e., location and scale) and the other determines what the object mask shape (and pose) should look like. The two modules are connected together via a spatial transformation network and jointly trained. We devise a learning procedure that leverage both supervised and unsupervised data and show our model can insert an object at diverse locations with various appearances. We conduct extensive experimental validations with comparisons to strong baselines to verify the effectiveness of the proposed network.