Computer Vision ¿ ECCV 2024

18th European Conference, Milan, Italy, September 29¿October 4, 2024, Proceedings, Part LVIII
Herausgegeben:Leonardis, Ales; Ricci, Elisa; Roth, Stefan; Russakovsky, Olga; Sattler, Torsten; Varol, Gül

• Broschiertes Buch

Produktbeschreibung

The multi-volume set of LNCS books with volume numbers 15059 up to 15147 constitutes the refereed proceedings of the 18th European Conference on Computer Vision, ECCV 2024, held in Milan, Italy, during September 29-October 4, 2024.

The 2387 papers presented in these proceedings were carefully reviewed and selected from a total of 8585 submissions. They 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; motion estimation.

Produktdetails

• Lecture Notes in Computer Science 15116
• Verlag: Springer / Springer Nature Switzerland / Springer, Berlin
• Artikelnr. des Verlages: 978-3-031-73635-3
• Seitenzahl: 572
• Erscheinungstermin: 5. November 2024
• Englisch
• Abmessung: 235mm x 155mm x 31mm
• Gewicht: 856g
• ISBN-13: 9783031736353
• ISBN-10: 3031736354
• Artikelnr.: 71495500

Inhaltsangabe

AWOL: Analysis WithOut synthesis using Language.- OneVOS: Unifying Video Object Segmentation with All-in-One Transformer Framework.- M3DBench: Towards Omni 3D Assistant with Interleaved Multi-modal Instructions.- MSD: A Benchmark Dataset for Floor Plan Generation of Building Complexes.- End-to-End Rate-Distortion Optimized 3D Gaussian Representation.- Temporal Residual Jacobians for Rig-free Motion Transfer.- LetsMap: Unsupervised Representation Learning for Label-Efficient Semantic BEV Mapping.- Deblurring 3D Gaussian Splatting.- Taming Lookup Tables for Efficient Image Retouching.- DualDn: Dual-domain Denoising via Differentiable ISP.- Quantization-Friendly Winograd Transformations for Convolutional Neural Networks.- A Task is Worth One Word: Learning with Task Prompts for High-Quality Versatile Image Inpainting.- Self-supervised Shape Completion via Involution and Implicit Correspondences.- From Fake to Real: Pretraining on Balanced Synthetic Images to Prevent Spurious Correlations in Image Recognition.- Cross-Domain Few-Shot Object Detection via Enhanced Open-Set Object Detector.- NICP: Neural ICP for 3D Human Registration at Scale.- PredBench: Benchmarking Spatio-Temporal Prediction across Diverse Disciplines.- FontStudio: Shape-Adaptive Diffusion Model for Coherent and Consistent Font Effect Generation.- Chronologically Accurate Retrieval for Temporal Grounding of Motion-Language Models.- StableDrag: Stable Dragging for Point-based Image Editing.- Improving Feature Stability during Upsampling -- Spectral Artifacts and the Importance of Spatial Context.- Dynamic Data Selection for Efficient SSL via Coarse-to-Fine Refinement.- Neural Surface Detection for Unsigned Distance. Fields.- One-Shot Diffusion Mimicker for Handwritten Text Generation.- Event-Based Motion Magnification.- Improving Neural Surface Reconstruction with Feature Priors from Multi-View Images.- Towards Multimodal Sentiment Analysis Debiasing via Bias Purification.