CVPR 2025 论文和开源项目合集 | CVer-Papers-with-Code

CVPR 2025 论文和开源项目合集(Papers with Code) 

CVPR 2025 decisions are now available on OpenReview！22.1% = 2878 / 13008

注1：欢迎各位大佬提交issue，分享CVPR 2025论文和开源项目！

注2：关于往年CV顶会论文以及其他优质CV论文和大盘点，详见： https://github.com/amusi/daily-paper-computer-vision

• ECCV 2024
• CVPR 2024

【CVPR 2025 论文开源目录】

• 3DGS(Gaussian Splatting)
• Avatars
• Backbone
• CLIP
• Mamba
• Embodied AI
• GAN
• GNN
• 多模态大语言模型(MLLM)
• 大语言模型(LLM)
• NAS
• OCR
• NeRF
• DETR
• 扩散模型(Diffusion Models)
• ReID(重识别)
• 长尾分布(Long-Tail)
• Vision Transformer
• 视觉和语言(Vision-Language)
• 自监督学习(Self-supervised Learning)
• 数据增强(Data Augmentation)
• 目标检测(Object Detection)
• 异常检测(Anomaly Detection)
• 目标跟踪(Visual Tracking)
• 语义分割(Semantic Segmentation)
• 实例分割(Instance Segmentation)
• 全景分割(Panoptic Segmentation)
• 医学图像(Medical Image)
• 医学图像分割(Medical Image Segmentation)
• 视频目标分割(Video Object Segmentation)
• 视频实例分割(Video Instance Segmentation)
• 参考图像分割(Referring Image Segmentation)
• 图像抠图(Image Matting)
• 图像编辑(Image Editing)
• Low-level Vision
• 超分辨率(Super-Resolution)
• 去噪(Denoising)
• 去模糊(Deblur)
• 自动驾驶(Autonomous Driving)
• 3D点云(3D Point Cloud)
• 3D目标检测(3D Object Detection)
• 3D语义分割(3D Semantic Segmentation)
• 3D目标跟踪(3D Object Tracking)
• 3D语义场景补全(3D Semantic Scene Completion)
• 3D配准(3D Registration)
• 3D人体姿态估计(3D Human Pose Estimation)
• 3D人体Mesh估计(3D Human Mesh Estimation)
• 医学图像(Medical Image)
• 图像生成(Image Generation)
• 视频生成(Video Generation)
• 3D生成(3D Generation)
• 视频理解(Video Understanding)
• 行为检测(Action Detection)
• 具身智能(Embodied AI)
• 文本检测(Text Detection)
• 知识蒸馏(Knowledge Distillation)
• 模型剪枝(Model Pruning)
• 图像压缩(Image Compression)
• 三维重建(3D Reconstruction)
• 深度估计(Depth Estimation)
• 轨迹预测(Trajectory Prediction)
• 车道线检测(Lane Detection)
• 图像描述(Image Captioning)
• 视觉问答(Visual Question Answering)
• 手语识别(Sign Language Recognition)
• 视频预测(Video Prediction)
• 新视点合成(Novel View Synthesis)
• Zero-Shot Learning(零样本学习)
• 立体匹配(Stereo Matching)
• 特征匹配(Feature Matching)
• 暗光图像增强(Low-light Image Enhancement)
• 场景图生成(Scene Graph Generation)
• 风格迁移(Style Transfer)
• 隐式神经表示(Implicit Neural Representations)
• 图像质量评价(Image Quality Assessment)
• 视频质量评价(Video Quality Assessment)
• 数据集(Datasets)
• 新任务(New Tasks)
• 其他(Others)

3DGS(Gaussian Splatting)

Avatars

Backbone

CLIP

Mamba

MambaVision: A Hybrid Mamba-Transformer Vision Backbone

• Paper: [2407.08083] MambaVision: A Hybrid Mamba-Transformer Vision Backbone
• Code: https://github.com/NVlabs/MambaVision

MobileMamba: Lightweight Multi-Receptive Visual Mamba Network

• Paper: [2411.15941] MobileMamba: Lightweight Multi-Receptive Visual Mamba Network
• Code: https://github.com/lewandofskee/MobileMamba

Embodied AI

CityWalker: Learning Embodied Urban Navigation from Web-Scale Videos

• Project: CityWalker
• Paper: [2411.17820] CityWalker: Learning Embodied Urban Navigation from Web-Scale Videos
• Code: https://github.com/ai4ce/CityWalker

GAN

OCR

NeRF

DETR

Prompt

多模态大语言模型(MLLM)

LSceneLLM: Enhancing Large 3D Scene Understanding Using Adaptive Visual Preferences

• Paper： [2412.01292] LSceneLLM: Enhancing Large 3D Scene Understanding Using Adaptive Visual Preferences
• Code: https://github.com/Hoyyyaard/LSceneLLM

DynRefer: Delving into Region-level Multimodal Tasks via Dynamic Resolution

• Paper: [2405.16071] DynRefer: Delving into Region-level Multimodal Tasks via Dynamic Resolution
• Code: https://github.com/callsys/DynRefer

Retrieval-Augmented Personalization for Multimodal Large Language Models

• Project Page: RAP-MLLM
• Paper: [2410.13360] Retrieval-Augmented Personalization for Multimodal Large Language Models
• Code: https://github.com/Hoar012/RAP-MLLM

大语言模型(LLM)

NAS

ReID(重识别)

From Poses to Identity: Training-Free Person Re-Identification via Feature Centralization

• Paper: [2503.00938] From Poses to Identity: Training-Free Person Re-Identification via Feature Centralization
• Code: https://github.com/yuanc3/Pose2ID

AirRoom: Objects Matter in Room Reidentification

• Project: https://sairlab.org/airroom/
• Paper: [2503.01130] AirRoom: Objects Matter in Room Reidentification

扩散模型(Diffusion Models)

TinyFusion: Diffusion Transformers Learned Shallow

• Paper: [2412.01199] TinyFusion: Diffusion Transformers Learned Shallow
• Code: https://github.com/VainF/TinyFusion

Vision Transformer

视觉和语言(Vision-Language)

NLPrompt: Noise-Label Prompt Learning for Vision-Language Models

• Paper: [2412.01256] NLPrompt: Noise-Label Prompt Learning for Vision-Language Models
• Code: https://github.com/qunovo/NLPrompt

目标检测(Object Detection)

LLMDet: Learning Strong Open-Vocabulary Object Detectors under the Supervision of Large Language Models

• Paper: [2501.18954] LLMDet: Learning Strong Open-Vocabulary Object Detectors under the Supervision of Large Language Models
• Code：https://github.com/iSEE-Laboratory/LLMDet

异常检测(Anomaly Detection)

目标跟踪(Object Tracking)

Multiple Object Tracking as ID Prediction

• Paper：[2403.16848] Multiple Object Tracking as ID Prediction
• Code: https://github.com/MCG-NJU/MOTIP

Omnidirectional Multi-Object Tracking

• Paper:[2503.04565] Omnidirectional Multi-Object Tracking
• Code:https://github.com/xifen523/OmniTrack

医学图像(Medical Image)

医学图像分割(Medical Image Segmentation)

自动驾驶(Autonomous Driving)

LiMoE: Mixture of LiDAR Representation Learners from Automotive Scenes

• Project: https://ldkong.com/LiMoE
• Paper: [2501.04004] LiMoE: Mixture of LiDAR Representation Learners from Automotive Scenes
• Code: https://github.com/Xiangxu-0103/LiMoE

3D点云(3D-Point-Cloud)

3D目标检测(3D Object Detection)

3D语义分割(3D Semantic Segmentation)

Low-level Vision

超分辨率(Super-Resolution)

AESOP: Auto-Encoded Supervision for Perceptual Image Super-Resolution

• Paper: [2412.00124] Auto-Encoded Supervision for Perceptual Image Super-Resolution
• Code: https://github.com/2minkyulee/AESOP-Auto-Encoded-Supervision-for-Perceptual-Image-Super-Resolution

去噪(Denoising)

图像去噪(Image Denoising)

3D人体姿态估计(3D Human Pose Estimation)

图像生成(Image Generation)

Reconstruction vs. Generation: Taming Optimization Dilemma in Latent Diffusion Models

• Paper: [2501.01423] Reconstruction vs. Generation: Taming Optimization Dilemma in Latent Diffusion Models
• Code: https://github.com/hustvl/LightningDiT

SleeperMark: Towards Robust Watermark against Fine-Tuning Text-to-image Diffusion Models

• Paper: [2412.04852] SleeperMark: Towards Robust Watermark against Fine-Tuning Text-to-image Diffusion Models
• Code: https://github.com/taco-group/SleeperMark

TokenFlow: Unified Image Tokenizer for Multimodal Understanding and Generation

• Homepage: TokenFlow
• Code: https://github.com/ByteFlow-AI/TokenFlow
• Paper:[2412.03069] TokenFlow: Unified Image Tokenizer for Multimodal Understanding and Generation

PAR: Parallelized Autoregressive Visual Generation

• Project: Parallelized Autoregressive Visual Generation
• Paper: [2412.15119] Parallelized Autoregressive Visual Generation
• Code: https://github.com/Epiphqny/PAR

Generative Photography: Scene-Consistent Camera Control for Realistic Text-to-Image Synthesis

• Project: Generative Photography: Scene-Consistent Camera Control for Realistic Text-to-Image Synthesis
• Paper: [2412.02168] Generative Photography: Scene-Consistent Camera Control for Realistic Text-to-Image Synthesis
• Code: https://github.com/pandayuanyu/generative-photography

视频生成(Video Generation)

Identity-Preserving Text-to-Video Generation by Frequency Decomposition

• Paper: [2411.17440] Identity-Preserving Text-to-Video Generation by Frequency Decomposition
• Code: https://github.com/PKU-YuanGroup/ConsisID

Cinemo: Consistent and Controllable Image Animation with Motion Diffusion Models

• Paper: [2407.15642] Cinemo: Consistent and Controllable Image Animation with Motion Diffusion Models
• Code: https://github.com/maxin-cn/Cinemo

X-Dyna: Expressive Dynamic Human Image Animation

• Paper: [2501.10021] X-Dyna: Expressive Dynamic Human Image Animation
• Code: https://github.com/bytedance/X-Dyna

PhyT2V: LLM-Guided Iterative Self-Refinement for Physics-Grounded Text-to-Video Generation

• Paper: https://arxiv.org/pdf/2412.00596
• Code: https://github.com/pittisl/PhyT2V

Timestep Embedding Tells: It's Time to Cache for Video Diffusion Model

• Project: Timestep Embedding Tells: It's Time to Cache for Video Diffusion Model
• Paper: [2411.19108] Timestep Embedding Tells: It's Time to Cache for Video Diffusion Model
• Code: https://github.com/ali-vilab/TeaCache

AR-Diffusion: Asynchronous Video Generation with Auto-Regressive Diffusion

• Project: AR-Diffusion: Asynchronous Video Generation with Auto-Regressive Diffusion
• Paper: [2503.07418] AR-Diffusion: Asynchronous Video Generation with Auto-Regressive Diffusion
• Code: https://github.com/iva-mzsun/AR-Diffusion

图像编辑(Image Editing)

Edit Away and My Face Will not Stay: Personal Biometric Defense against Malicious Generative Editing

• Paper: [2411.16832] Edit Away and My Face Will not Stay: Personal Biometric Defense against Malicious Generative Editing
• Code: https://github.com/taco-group/FaceLock

h-Edit: Effective and Flexible Diffusion-Based Editing via Doob’s h-Transform

• Paper: [2503.02187] h-Edit: Effective and Flexible Diffusion-Based Editing via Doob's h-Transform
• Code: https://github.com/nktoan/h-edit

视频编辑(Video Editing)

3D生成(3D Generation)

Generative Gaussian Splatting for Unbounded 3D City Generation

• Project: GaussianCity | Infinite Script
• Paper: [2406.06526] Generative Gaussian Splatting for Unbounded 3D City Generation
• Code: https://github.com/hzxie/GaussianCity

StdGEN: Semantic-Decomposed 3D Character Generation from Single Images

• Project: StdGEN
• Paper: [2411.05738] StdGEN: Semantic-Decomposed 3D Character Generation from Single Images
• Code: https://github.com/hyz317/StdGEN

3D重建(3D Reconstruction)

Fast3R: Towards 3D Reconstruction of 1000+ Images in One Forward Pass

• Project: Fast3R: Towards 3D Reconstruction of 1000+ Images in One Forward Pass
• Paper: [2501.13928] Fast3R: Towards 3D Reconstruction of 1000+ Images in One Forward Pass

人体运动生成(Human Motion Generation)

SemGeoMo: Dynamic Contextual Human Motion Generation with Semantic and Geometric Guidance

• Project: SemGeoMo: Dynamic Contextual Human Motion Generation with Semantic and Geometric Guidance
• Paper: [2503.01291] SemGeoMo: Dynamic Contextual Human Motion Generation with Semantic and Geometric Guidance
• https://github.com/4DVLab/SemGeoMo

视频理解(Video Understanding)

Temporal Grounding Videos like Flipping Manga

• Paper: [2411.10332] Number it: Temporal Grounding Videos like Flipping Manga
• Code: https://github.com/yongliang-wu/NumPro

具身智能(Embodied AI)

Universal Actions for Enhanced Embodied Foundation Models

• Project: Universal Actions for Enhanced Embodied Foundation Models
• Paper: [2501.10105] Universal Actions for Enhanced Embodied Foundation Models
• Code: https://github.com/2toinf/UniAct

知识蒸馏(Knowledge Distillation)

深度估计(Depth Estimation)

DepthCrafter: Generating Consistent Long Depth Sequences for Open-world Videos

• Project: https://depthcrafter.github.io
• Paper: [2409.02095] DepthCrafter: Generating Consistent Long Depth Sequences for Open-world Videos
• Code: https://github.com/Tencent/DepthCrafter

MonSter: Marry Monodepth to Stereo Unleashes Power

• Paper: [2501.08643] MonSter: Marry Monodepth to Stereo Unleashes Power
• Code: https://github.com/Junda24/MonSter

立体匹配(Stereo Matching)

MonSter: Marry Monodepth to Stereo Unleashes Power

• Paper: [2501.08643] MonSter: Marry Monodepth to Stereo Unleashes Power
• Code: https://github.com/Junda24/MonSter

暗光图像增强(Low-light Image Enhancement)

HVI: A New color space for Low-light Image Enhancement

• Paper: [2502.20272] HVI: A New Color Space for Low-light Image Enhancement
• Code: https://github.com/Fediory/HVI-CIDNet
• Demo: https://huggingface.co/spaces/Fediory/HVI-CIDNet_Low-light-Image-Enhancement_

ReDDiT: Efficient Diffusion as Low Light Enhancer

• Paper: [2410.12346] Efficient Diffusion as Low Light Enhancer
• Code: https://github.com/lgz-0713/ReDDiT

场景图生成(Scene Graph Generation)

风格迁移(Style Transfer)

StyleStudio: Text-Driven Style Transfer with Selective Control of Style Elements

• Project: StyleStudio: Text-Driven Style Transfer with Selective Control of Style Elements
• Paper: [2412.08503] StyleStudio: Text-Driven Style Transfer with Selective Control of Style Elements
• Code: https://github.com/Westlake-AGI-Lab/StyleStudio

视频质量评价(Video Quality Assessment)

数据集(Datasets)

其他(Others)

注：All copied from [GitHub - amusi/CVPR2025-Papers-with-Code: CVPR 2025 论文和开源项目合集]