Computer Vision – ECCV 2024

Diffusion Model is a Good Pose Estimator from 3D RF-Vision.- UPose3D: Uncertainty-Aware 3D Human Pose Estimation with Cross-View and Temporal Cues.- Learning 3D-aware GANs from Unposed Images with Template Feature Field.- TAPTR: Tracking Any Point with Transformers as Detection.- Token Compensator: Altering Inference Cost of Vision Transformer without Re-Tuning.- Point-supervised Panoptic Segmentation via Estimating Pseudo Labels from Learnable Distance.- BRAVE: Broadening the visual encoding of vision-language models.- HUMOS: Human Motion Model Conditioned on Body Shape.- Omni-Recon: Harnessing Image-based Rendering for General-Purpose Neural Radiance Fields.- MVDiffHD: A Dense High-resolution Multi-view Diffusion Model for Single or Sparse-view 3D Object Reconstruction.- FlowCon: Out-of-Distribution Detection using Flow-based Contrastive Learning.- LEIA: Latent View-invariant Embeddings for Implicit 3D Articulation.- Un-EVIMO: Unsupervised Event-based Independent Motion Segmentation.- Seeing the Unseen: A Frequency Prompt Guided Transformer for Image Restoration.- CityGaussian: Real-time High-quality Large-Scale Scene Rendering with Gaussians.- Bayesian Evidential Deep Learning for Online Action Detection.- AdaNAT: Exploring Adaptive Policy for Token-Based Image Generation.- Rethinking Data Augmentation for Robust LiDAR Semantic Segmentation in Adverse Weather.- Diffusion-Generated Pseudo-Observations for High-Quality Sparse-View Reconstruction.- Memory-Efficient Fine-Tuning for Quantized Diffusion Model.- VCD-Texture: Variance Alignment based 3D-2D Co-Denoising for Text-Guided Texturing.- MotionLCM: Real-time Controllable Motion Generation via Latent Consistency Model.- Human Hair Reconstruction with Strand-Aligned 3D Gaussians.- COIN: Control-Inpainting Diffusion Prior for Human and Camera Motion Estimation.- SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders.- Bridge Past and Future: Overcoming Information Asymmetry in Incremental Object Detection.- Global-to-Pixel Regression for Human Mesh Recovery.