Computer Vision - Eccv 2022: 17th European Conference, Tel Aviv, Israel, October 23-27, 2022, Proceedings, Part XIV

ISBN-13: 9783031197802 / Angielski / Miękka / 2022 / 751 str.

Shai Avidan; Gabriel Brostow; Moustapha Cissé

Opis

The 39-volume set, comprising the LNCS books 13661 until 13699, constitutes the refereed proceedings of the 17th European Conference on Computer Vision, ECCV 2022, held in Tel Aviv, Israel, during October 23–27, 2022. The 1645 papers presented in these proceedings were carefully reviewed and selected from a total of 5804 submissions. The papers deal with topics such as computer vision; machine learning; deep neural networks; reinforcement learning; object recognition; image classification; image processing; object detection; semantic segmentation; human pose estimation; 3d reconstruction; stereo vision; computational photography; neural networks; image coding; image reconstruction; object recognition; motion estimation.

Spis treści

Watermark Vaccine: Adversarial Attacks to Prevent Watermark Removal.- Explaining Deepfake Detection by Analysing Image Matching.- FrequencyLowCut Pooling – Plug & Play against Catastrophic Overfitting.- TAFIM: Targeted Adversarial Attacks against Facial Image Manipulations.- FingerprintNet: Synthesized Fingerprints for Generated Image Detection.- Detecting Generated Images by Real Images.- An Information Theoretic Approach for Attention-Driven Face Forgery Detection.- Exploring Disentangled Content Information for Face Forgery Detection.- RepMix: Representation Mixing for Robust Attribution of Synthesized Images.- Totems: Physical Objects for Verifying Visual Integrity.- Dual-Stream Knowledge-Preserving Hashing for Unsupervised Video Retrieval.- PASS: Part-Aware Self-Supervised Pre-training for Person Re-identification.- Adaptive Cross-Domain Learning for Generalizable Person Re-identification.- Multi-Query Video Retrieval.- Hierarchical Average Precision Training for Pertinent Image Retrieval.- Learning Semantic Correspondence with Sparse Annotations.- Dynamically Transformed Instance Normalization Network for Generalizable Person Re-identification.- Domain Adaptive Person Search.- TS2-Net: Token Shift and Selection Transformer for Text-Video Retrieval.- Unstructured Feature Decoupling for Vehicle Re-identification.- Deep Hash Distillation for Image Retrieval.- Mimic Embedding via Adaptive Aggregation: Learning Generalizable Person Re-identification.- Granularity-Aware Adaptation for Image Retrieval over Multiple Tasks.- Learning Audio-Video Modalities from Image Captions.- RVSL: Robust Vehicle Similarity Learning in Real Hazy Scenes Based on Semi-Supervised Learning.- Lightweight Attentional Feature Fusion: A New Baseline for Text-to-Video Retrieval.- Modality Synergy Complement Learning with Cascaded Aggregation for Visible-Infrared Person Re-identification.- Cross-Modality Transformer for Visible-Infrared Person Re-identification.- Audio-Visual Mismatch-Aware Video Retrieval via Association and Adjustment.- Connecting Compression Spaces with Transformer for Approximate Nearest Neighbor Search.- SEMICON: A Learning-to-Hash Solution for Large-Scale Fine-Grained Image Retrieval.- CAViT: Contextual Alignment Vision Transformer for Video Object Re-identification.- Text-Based Temporal Localization of Novel Events.- Reliability-Aware Prediction via Uncertainty Learning for Person Image Retrieval.- Relighting4D: Neural Relightable Human from Videos.- Real-Time Intermediate Flow Estimation for Video Frame Interpolation.- PixelFolder: An Efficient Progressive Pixel Synthesis Network for Image Generation.- StyleSwap: Style-Based Generator Empowers Robust Face Swapping.- Paint2Pix: Interactive Painting Based Progressive Image Synthesis and Editing.- FurryGAN: High Quality Foreground-Aware Image Synthesis.- SCAM! Transferring Humans between Images with Semantic Cross

Attention Modulation.- Sem2NeRF: Converting Single-View Semantic Masks to Neural

Radiance Fields.