CVPR-2024-Oral

Geometry

Rethinking Inductive Biases for Surface Normal Estimation

• Homepage : https://baegwangbin.github.io/DSINE/
• Paper : https://arxiv.org/abs/2403.00712

Scene Understanding

PaSCo: Urban 3D Panoptic Scene Completion with Uncertainty Awareness

• Homepage : https://astra-vision.github.io/PaSCo/
• Paper : https://arxiv.org/abs/2312.02158

SceneFun3D: Fine-Grained Functionality and Affordance Understanding in 3D Scenes

• Homepage : https://scenefun3d.github.io/
• Paper : https://alexdelitzas.github.io/assets/pdf/SceneFun3D.pdf

Reconstruction

pixelSplat: 3D Gaussian Splats from Image Pairs for Scalable Generalizable 3D Reconstruction

• Homepage : https://davidcharatan.com/pixelsplat/
• Paper : https://arxiv.org/abs/2312.12337

Embodied AI

Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectives

• Homepage : https://ego-exo4d-data.org/
• Paper : https://arxiv.org/abs/2311.18259
• Description : Dataset, Benchmark

Multi-Modal

Modeling Multimodal Social Interactions: New Challenges and Baselines with Densely Aligned Representations

• Paper : https://arxiv.org/abs/2403.02090
• Descritpion : Baseline

Video

FMA-Net: Flow-Guided Dynamic Filtering and Iterative Feature Refinement with Multi-Attention for Joint Video Super-Resolution and Deblurring

• Homepage : https://kaist-viclab.github.io/fmanet-site/
• Paper : https://arxiv.org/abs/2401.03707
• Description : Method

CVPR-2024-Highlight

• 3DGS
• Diffusion
• Presentation
• LLM
• 3D Generation
• 2D Generation
• Digital Human
• Human-Scene
• Multi-Modal
• Video
• Image
• Novel View Synthesis
• Reconstruction
• Segmentation
• Embodied AI
• Pose Estimation
• Autonomous Driving
• SLAM
• Medical Image Analysis
• Machine Learning
• Robotic Manipulation
• Others

3D Presentation(GS, NeRF)

HybridNeRF: Efficient Neural Rendering via Adaptive Volumetric Surfaces

• Homepage : https://haithemturki.com/hybrid-nerf/
• Paper : https://haithemturki.com/hybrid-nerf/resources/paper.pdf
• Description : Presentation

HashPoint: Accelerated Point Searching and Sampling for Neural Rendering

• Homepage : https://jiahao-ma.github.io/hashpoint/
• Paper : https://arxiv.org/abs/2404.14044v1
• Description : Method Theory

Scaffold-GS: Structured 3D Gaussians for View-Adaptive Rendering

• Homepage : https://city-super.github.io/scaffold-gs/
• Paper: https://arxiv.org/abs/2312.00109

Supercharging 3D Gaussian Splatting to Enable Distilled Feature Fields

• Homepage : https://feature-3dgs.github.io/
• Paper : https://arxiv.org/abs/2312.03203

Diffusion

DistriFusion: Distributed Parallel Inference for High-Resolution Diffusion Models

• Homepage : https://migcproject.github.io/
• Paper : https://arxiv.org/abs/2402.19481

Style Injection in Diffusion: A Training-free Approach for Adapting Large-scale Diffusion Models for Style Transfer

• Homepage : https://jiwoogit.github.io/StyleID_site/
• Paper : https://arxiv.org/abs/2312.09008

TFMQ-DM: Temporal Feature Maintenance Quantization for Diffusion Models

• Paper : https://arxiv.org/abs/2311.16503

Presentation

Chat-UniVi: Unified Visual Representation Empowers Large Language Models with Image and Video Understanding

• Paper : https://arxiv.org/abs/2311.08046

LLM

VCD: Mitigating Object Hallucinations in Large Vision-Language Models through Visual Contrastive Decoding

• Paper : https://arxiv.org/abs/2311.16922

3D Generation

RichDreamer: A Generalizable Normal-Depth Diffusion Model for Detail Richness in Text-to-3D

• Homepage : https://aigc3d.github.io/richdreamer/
• Paper : https://arxiv.org/abs/2311.16918
• Description : Method

MeshGPT: Generating Triangle Meshes with Decoder-Only Transformers

• Homepage : https://nihalsid.github.io/mesh-gpt/
• Paper : https://arxiv.org/abs/2311.15475
• Description : Method, Mesh Generation

PhysGaussian: Physics-Integrated 3D Gaussians for Generative Dynamics

• Homepage : https://xpandora.github.io/PhysGaussian/
• Paper : https://arxiv.org/abs/2311.12198

SceneTex: High-Quality Texture Synthesis for Indoor Scenes via Diffusion Priors

• Homepage : https://daveredrum.github.io/SceneTex/
• Paper : https://arxiv.org/abs/2311.17261

2D Generation

MIGC: Multi-Instance Generation Controller for Text-to-Image Synthesis

• Homepage : https://migcproject.github.io/
• Paper : https://arxiv.org/pdf/2402.05408

OpenBias: Open-set Bias Detection in Text-to-Image Generative Models

• Paper : https://arxiv.org/abs/2404.07990

Digital Human

A 4D Dataset ofReal-World Human Clothing With Semantic Annotations

• Homepage : https://eth-ait.github.io/4d-dress/
• Paper : https://arxiv.org/abs/2404.18630
• Description : Dataset

SIFU: Side-view Conditioned Implicit Function for Real-world Usable Clothed Human Reconstruction

• Homepage : https://river-zhang.github.io/SIFU-projectpage/
• Paper : https://arxiv.org/abs/2312.06704

3D Human Pose Perception from Egocentric Stereo Videos

• Homepage : https://4dqv.mpi-inf.mpg.de/UnrealEgo2/
• Paper : https://arxiv.org/abs/2401.00889
• Description : Benchmark Dataset

Relightable and Animatable Neural Avatar from Sparse-View Video

• Homepage : https://zju3dv.github.io/relightable_avatar/
• Paper : https://arxiv.org/abs/2308.07903

HumanGaussian: Text-Driven 3D Human Generation with Gaussian Splatting

• Homepage : https://alvinliu0.github.io/projects/HumanGaussian
• Paper : https://arxiv.org/abs/2311.17061

GaussianAvatars: Photorealistic Head Avatars with Rigged 3D Gaussians

• Homepage : https://github.com/ShenhanQian/GaussianAvatars

Generalizable Pixel-wise 3D Gaussian Splatting for Real-time Human Novel View Synthesis

• Homepage : https://shunyuanzheng.github.io/GPS-Gaussian
• Paper : https://arxiv.org/pdf/2312.02155

Human-Scene

Scaling Up Dynamic Human-Scene Interaction Modeling

• Homepage : https://jnnan.github.io/trumans/
• Paper : https://arxiv.org/abs/2403.08629
• Description : Dataset

Move as You Say, Interact as You Can: Language-guided Human Motion Generation with Scene Affordance

• Homepage : https://afford-motion.github.io/
• Paper : https://arxiv.org/abs/2403.18036
• Description :

Multi-Modal

OPERA: Alleviating Hallucination in Multi-Modal Large Language Models via Over-Trust Penalty and Retrospection-Allocation

• Homepage : https://github.com/shikiw/OPERA
• Paper : https://arxiv.org/abs/2311.17911
• Description : Method

Prompt Highlighter: Interactive Control for Multi-Modal LLMs

• Homepage : https://julianjuaner.github.io/projects/PromptHighlighter/
• Paper : https://arxiv.org/abs/2312.04302

Cooperation Does Matter: Exploring Multi-Order Bilateral Relations for Audio-Visual Segmentation

• Paper: https://arxiv.org/pdf/2312.06462

Monkey: Image Resolution and Text Label Are Important Things for Large Multi-modal Models

• Paper : https://arxiv.org/abs/2311.06607

Video

3DGStream: On-the-Fly Training of 3D Gaussians for Efficient Streaming of Photo-Realistic Free-Viewpoint Videos

• Homepage : https://sjojok.github.io/3dgstream/
• Paper : https://arxiv.org/abs/2403.01444

SpatialTracker: Tracking Any 2D Pixels in 3D Space

• Homepage : https://henry123-boy.github.io/SpaTracker/
• Paper : https://arxiv.org/abs/2404.04319

Upscale-A-Video: Temporal-Consistent Diffusion Model for Real-World Video Super-Resolution

• Homepage : https://shangchenzhou.com/projects/upscale-a-video/
• Paper : https://arxiv.org/abs/2312.06640

RAVE: Randomized Noise Shuffling for Fast and Consistent Video Editing with Diffusion Models

• Homepage : https://rave-video.github.io/
• Paper : https://arxiv.org/abs/2312.04524

CoDeF: Content Deformation Fields for Temporally Consistent Video Processing

• Homepage : https://qiuyu96.github.io/CoDeF/
• Paper : https://arxiv.org/abs/2308.07926

Putting the Object Back into Video Object Segmentation

• Homepage : https://hkchengrex.com/Cutie/
• Paper : https://arxiv.org/abs/2310.12982

Boosting Neural Representations for Videos with a Conditional Decoder

• Paper : https://arxiv.org/abs/2402.18152

Enhancing Video Super-Resolution via Implicit Resampling-based Alignment

• Paper : https://arxiv.org/abs/2305.00163

VTimeLLM: Empower LLM to Grasp Video Moments

• Homepage : https://arxiv.org/abs/2311.18445

Bridging the Gap: A Unified Video Comprehension Framework for Moment Retrieval and Highlight Detection

• Paper : https://arxiv.org/abs/2311.16464

Image

ImageNet-D: Benchmarking Neural Network Robustness on Diffusion Synthetic Object

• Homepage : https://chenshuang-zhang.github.io/imagenet_d/
• Paper : https://arxiv.org/abs/2403.18775
• Description : Benchmark Dataset

SmartEdit: Exploring Complex Instruction-based Image Editing with Multimodal Large Language Models

• Homepage : https://yuzhou914.github.io/SmartEdit/
• Paper : https://arxiv.org/abs/2312.06739

Coarse-to-Fine Latent Diffusion for Pose-Guided Person Image Synthesis

• Paper : https://arxiv.org/abs/2402.18078

Novel View Synthesis

CoPoNeRF: Unifying Correspondence, Pose and NeRF for Pose-Free Novel View Synthesis from Stereo Pairs

• Homepage : https://ku-cvlab.github.io/CoPoNeRF/
• Paper : https://arxiv.org/abs/2312.07246

SCINeRF: Neural Radiance Fields from a Snapshot Compressive Image

• Paper : https://arxiv.org/abs/2403.20018

Reconstruction

IPoD: Implicit Field Learning with Point Diffusion for Generalizable 3D Object Reconstruction from Single RGB-D Images

• Homepage : https://yushuang-wu.github.io/IPoD/
• Paper : https://arxiv.org/abs/2404.00269

Gaussian-Flow: 4D Reconstruction with Dynamic 3D Gaussian Particle

• Homepage : https://nju-3dv.github.io/projects/Gaussian-Flow/
• Paper : https://arxiv.org/abs/2312.03431

Living Scenes: Multi-object Relocalization and Reconstruction in Changing 3D Environments

• Homepage : https://www.zhuliyuan.net/livingscenes
• Paper : https://arxiv.org/abs/2312.09138

HOLD: Category-agnostic 3D Reconstruction of Interacting Hands and Objects from Video

• Homepage : https://zc-alexfan.github.io/hold
• Paper : https://arxiv.org/abs/2311.18448

Segmentation

EAGLE: Eigen Aggregation Learning for Object-Centric Unsupervised Semantic Segmentation

• Homepage : https://micv-yonsei.github.io/eagle2024/
• Paper : https://arxiv.org/abs/2403.01482

GraCo: Granularity-Controllable Interactive Segmentation

• Homepage : https://zhao-yian.github.io/GraCo/
• Paper : https://arxiv.org/abs/2405.00587

OpenESS: Event-Based Semantic Scene Understanding with Open Vocabularies

• Paper : https://arxiv.org/abs/2405.05259

Frequency-Adaptive Dilated Convolution for Semantic Segmentation

• Paper : https://arxiv.org/abs/2403.05369

SPOT: Self-Training with Patch-Order Permutation for Object-Centric Learning with Autoregressive Transformers

• Paper : https://arxiv.org/abs/2312.00648

Embodied AI

PhyScene: Physically Interactable 3D Scene Synthesis for Embodied AI

• Homepage : https://physcene.github.io/
• Paper : https://arxiv.org/abs/2404.09465.pdf
• Description : Method

Pose Estimation

FoundationPose: Unified 6D Pose Estimation and Tracking of Novel Objects

• Homepage : https://nvlabs.github.io/FoundationPose/
• Paper : https://arxiv.org/abs/2312.08344

Autonomous Driving

Visual Point Cloud Forecasting enables Scalable Autonomous Driving

• Paper : https://arxiv.org/abs/2312.17655

Generalized Predictive Model for Autonomous Driving

• Paper : https://arxiv.org/abs/2403.09630

Dynamic LiDAR Re-simulation using Compositional Neural Fields

• Paper : https://arxiv.org/abs/2312.05247

SLAM

Gaussian Splatting SLAM

• Homepage : https://rmurai.co.uk/projects/GaussianSplattingSLAM/
• Paper : https://arxiv.org/abs/2312.06741
• Description : Method

SFM

VGGSfM: Visual Geometry Grounded Deep Structure From Motion

• Homepage : https://vggsfm.github.io/
• Paper : https://arxiv.org/abs/2312.04563

Camera Pose

Map-Relative Pose Regression for Visual Re-Localization

• Homepage : https://nianticlabs.github.io/marepo/
• Paper : https://arxiv.org/abs/2404.09884

FAR: Flexible, Accurate and Robust 6DoF Relative Camera Pose Estimation

• Paper : https://arxiv.org/abs/2311.17597

Medical Image Analysis

Continual Self-supervised Learning: Towards Universal Multi-modal Medical Data Representation Learning

• Paper : https://arxiv.org/abs/2311.17597

Diversified and Personalized Multi-rater Medical Image Segmentation

• Paper : https://arxiv.org/abs/2403.13417

Robotic Manipulation

Diffusion-EDFs: Bi-equivariant Denoising Generative Modeling on SE(3) for Visual Robotic Manipulation

• Homepage : https://sites.google.com/view/diffusion-edfs
• Paper : https://arxiv.org/abs/2309.02685

Knowledge Distillation

Logit Standardization in Knowledge Distillation

• Homepage : https://sunsean21.github.io/logit-stand-KD.html
• Paper : https://arxiv.org/abs/2403.01427

Feature Matching

Efficient LoFTR: Semi-Dense Local Feature Matching with Sparse-Like Speed

• Homepage : https://zju3dv.github.io/efficientloftr/
• Paper : https://zju3dv.github.io/efficientloftr/files/EfficientLoFTR.pdf

Stereo Matching

Selective-Stereo: Adaptive Frequency Information Selection for Stereo Matching

• Paper : https://arxiv.org/abs/2403.00486

Brain Decoding

MindBridge: A Cross-Subject Brain Decoding Framework

• Homepage : https://littlepure2333.github.io/MindBridge/
• Paper : https://arxiv.org/abs/2404.07850