Synthesizing Photorealistic Images with Deep Generative Learning

Abstract

The goal of this thesis is to present my research contributions towards solving various visual synthesis and generation tasks, comprising image translation, image completion, and completed scene decomposition. This thesis consists of five pieces of work, each of which presents a new learning-based approach for synthesizing images with plausible content as well as visually realistic appearance. Each work demonstrates the superiority of the proposed approach on image synthesis, with some further contributing to other tasks, such as depth estimation.

PhD thesis

References in corpus (27)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Conditional Generative Adversarial Nets
• GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
• Fully Convolutional Networks for Semantic Segmentation
• Unsupervised Domain Adaptation by Backpropagation
• Depth Map Prediction from a Single Image using a Multi-Scale Deep Network
• NICE: Non-linear Independent Components Estimation
• Learning Representations by Maximizing Mutual Information Across Views
• CyCADA: Cycle-Consistent Adversarial Domain Adaptation
• Depth Extraction from Video Using Non-parametric Sampling
• Variational Approaches for Auto-Encoding Generative Adversarial Networks
• Hybrid Task Cascade for Instance Segmentation
• Contrastive Learning for Unpaired Image-to-Image Translation
• Photographic Image Synthesis with Cascaded Refinement Networks
• One-Sided Unsupervised Domain Mapping
• Contextual Residual Aggregation for Ultra High-Resolution Image Inpainting
• High Resolution Face Completion with Multiple Controllable Attributes via Fully End-to-End Progressive Generative Adversarial Networks
• Rethinking Image Inpainting via a Mutual Encoder-Decoder with Feature Equalizations
• UnrealCV: Connecting Computer Vision to Unreal Engine
• Geometry-Aware Symmetric Domain Adaptation for Monocular Depth Estimation
• The Spatially-Correlative Loss for Various Image Translation Tasks
• High-Resolution Image Inpainting with Iterative Confidence Feedback and Guided Upsampling
• TSIT: A Simple and Versatile Framework for Image-to-Image Translation
• Visualizing the Invisible: Occluded Vehicle Segmentation and Recovery
• Learning Object-Compositional Neural Radiance Field for Editable Scene Rendering
• Spherical Image Generation from a Single Normal Field of View Image by Considering Scene Symmetry
• Visiting the Invisible: Layer-by-Layer Completed Scene Decomposition