Computer Vision - ECCV 2022

Editing Out-of-Domain GAN Inversion via Differential Activations.- On the Robustness of Quality Measures for GANs.- Sound-Guided Semantic Video Generation.- Inpainting at Modern Camera Resolution by Guided PatchMatch with Auto-Curation.- Controllable Video Generation through Global and Local Motion Dynamics.- StyleHEAT: One-Shot High-Resolution Editable Talking Face Generation via Pre-trained StyleGAN.- Long Video Generation with Time-Agnostic VQGAN and Time-Sensitive Transformer.- Combining Internal and External Constraints for Unrolling Shutter in Videos.- WISE: Whitebox Image Stylization by Example-Based Learning.- Neural Radiance Transfer Fields for Relightable Novel-View Synthesis with Global Illumination.- Transformers As Meta-Learners for Implicit Neural Representations.- Style Your Hair: Latent Optimization for Pose-Invariant Hairstyle Transfer via Local-Style-Aware Hair Alignment.- High-Resolution Virtual Try-On withMisalignment and OcclusionHandled Conditions.- A Codec Information Assisted Framework for Efficient Compressed Video Super-Resolution.- Injecting 3D Perception of Controllable NeRF-GAN into StyleGAN for Editable Portrait Image Synthesis.- AdaNeRF: Adaptive Sampling for Real-Time Rendering of Neural Radiance Fields.- Improving the Perceptual Quality of 2D Animation Interpolation.- Selective TransHDR: Transformer-Based Selective HDR Imaging Using Ghost Region Mask.- Learning Series-Parallel Lookup Tables for Efficient Image Super-Resolution.- GeoAug: Data Augmentation for Few-Shot NeRF with Geometry Constraints.- DoodleFormer: Creative Sketch Drawing with Transformers.- Implicit Neural Representations for Variable Length Human Motion Generation.- Learning Object Placement via Dual-Path Graph Completion.- Expanded Adaptive Scaling Normalization for End to End Image Compression.- Generator Knows What Discriminator Should Learn