CVPR'23 最新 70 篇论文分方向整理｜包含目标检测、图像处理、人脸、医学影像、半监督
编辑丨极市平台
CVPR2023已经放榜，今年有2360篇，接收率为25.78%。在CVPR2023正式会议召开前，为了让大家更快地获取和学习到计算机视觉前沿技术，极市对CVPR023 最新论文进行追踪，包括分研究方向的论文、代码汇总以及论文技术直播分享。
CVPR 2023 论文分方向整理目前在极市社区持续更新中，已累计更新了158篇，项目地址：https://www.cvmart.net/community/detail/7422
以下是最近更新的 CVPR 2023 论文，包含目标检测、图像处理、人脸、场景重建、医学影像、半监督学习/弱监督学习/无监督学习/自监督学习等方向。
可打包下载：https://www.cvmart.net/community/detail/7429
检测
2D目标检测(2D Object Detection
[1]CapDet: Unifying Dense Captioning and Open-World Detection Pretraining
paper：https://arxiv.org/abs/2303.02489
[2]Enhanced Training of Query-Based Object Detection via Selective Query Recollection
paper：https://arxiv.org/abs/2212.07593
code：https://github.com/Fangyi-Chen/SQR
[3]DETRs with Hybrid Matching
paper：https://arxiv.org/abs/2207.13080
code：https://github.com/HDETR
[4]YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors(YOLOv7
paper：https://arxiv.org/abs/2207.02696
code：https://github.com/WongKinYiu/yolov7
视频目标检测(Video Object Detection
[1]SCOTCH and SODA: A Transformer Video Shadow Detection Framework
paper：https://arxiv.org/abs/2211.06885
3D目标检测(3D object detection
[1]MSMDFusion: Fusing LiDAR and Camera at Multiple Scales with Multi-Depth Seeds for 3D Object Detection
paper：https://arxiv.org/abs/2209.03102
code：https://github.com/sxjyjay/msmdfusion
[2]Uni3D: A Unified Baseline for Multi-dataset 3D Object Detection
paper：https://arxiv.org/abs/2303.06880
code：https://github.com/PJLab-ADG/3DTrans
[3]LoGoNet: Towards Accurate 3D Object Detection with Local-to-Global Cross-Modal Fusion
paper：https://arxiv.org/abs/2303.03595
code：https://github.com/sankin97/LoGoNet
[4]ConQueR: Query Contrast Voxel-DETR for 3D Object Detection(3D 目标检测的Query Contrast Voxel-DETR
paper：https://arxiv.org/abs/2212.07289
code：https://github.com/poodarchu/ConQueR
显著性目标检测(Saliency Object Detection
[1]Texture-guided Saliency Distilling for Unsupervised Salient Object Detection
paper：https://arxiv.org/abs/2207.05921
code：https://github.com/moothes/A2S-v2
车道线检测(Lane Detection
[1]BEV-LaneDet: a Simple and Effective 3D Lane Detection Baseline
paper：https://arxiv.org/abs/2210.06006
异常检测(Anomaly Detection
[1]Block Selection Method for Using Feature Norm in Out-of-distribution Detection
paper：https://arxiv.org/abs/2212.02295
[2]Lossy Compression for Robust Unsupervised Time-Series Anomaly Detection
paper：https://arxiv.org/abs/2212.02303
[3]Multimodal Industrial Anomaly Detection via Hybrid Fusion
paper：https://arxiv.org/abs/2303.00601
code：https://github.com/nomewang/M3DM
分割(Segmentation
图像分割(Image Segmentation
[1]MP-Former: Mask-Piloted Transformer for Image Segmentation
paper：https://arxiv.org/abs/2303.07336
code：https://github.com/IDEA-Research/MP-Former
[2]Interactive Segmentation as Gaussian Process Classification
paper：https://arxiv.org/abs/2302.14578
语义分割(Semantic Segmentation
[1]Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP
paper：http://arxiv.org/abs/2210.04150
code：https://github.com/facebookresearch/ov-seg
[2]Efficient Semantic Segmentation by Altering Resolutions for Compressed Videos
paper：https://arxiv.org/abs/2303.07224
code：https://github.com/THU-LYJ-Lab/AR-Seg
[3]SCPNet: Semantic Scene Completion on Point Cloud
paper：https://arxiv.org/abs/2303.06884
[4]On Calibrating Semantic Segmentation Models: Analyses and An Algorithm
paper：https://arxiv.org/abs/2212.12053
[5]Learning Open-vocabulary Semantic Segmentation Models From Natural Language Supervision
paper：https://arxiv.org/abs/2301.09121
[6]Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic Segmentation
paper：https://arxiv.org/abs/2208.09910
code：https://github.com/LiheYoung/UniMatch
[7]Foundation Model Drives Weakly Incremental Learning for Semantic Segmentation
paper：https://arxiv.org/abs/2302.14250
实例分割(Instance Segmentation
[1]ISBNet: a 3D Point Cloud Instance Segmentation Network with Instance-aware Sampling and Box-aware Dynamic Convolution
paper：https://arxiv.org/abs/2303.00246
[22]PolyFormer: Referring Image Segmentation as Sequential Polygon Generation(PolyFormer：将图像分割表述为顺序多边形生成
paper：https://arxiv.org/abs/2302.07387
目标跟踪(Object Tracking
[1]Observation-Centric SORT: Rethinking SORT for Robust Multi-Object Tracking
paper：https://arxiv.org/abs/2203.14360v2
code：https://github.com/noahcao/OC_SORT
[2]Focus On Details: Online Multi-object Tracking with Diverse Fine-grained Representation
paper：https://arxiv.org/abs/2302.14589
[3]Referring Multi-Object Tracking
paper：https://arxiv.org/abs/2303.03366
[4]Simple Cues Lead to a Strong Multi-Object Tracker
paper：https://arxiv.org/abs/2206.04656
图像处理(Image Processing
超分辨率(Super Resolution
[1]Denoising Diffusion Probabilistic Models for Robust Image Super-Resolution in the Wild(野外鲁棒图像超分辨率的去噪扩散概率模型
paper：https://arxiv.org/abs/2302.07864
project：https://sihyun.me/PVDM/
图像复原/图像增强/图像重建(Image Restoration/Image Reconstruction
[1]Learning Distortion Invariant Representation for Image Restoration from A Causality Perspective
paper：https://arxiv.org/abs/2303.06859
code：https://github.com/lixinustc/Casual-IRDIL
[2]DR2: Diffusion-based Robust Degradation Remover for Blind Face Restoration
paper：https://arxiv.org/abs/2303.06885
[3]Robust Unsupervised StyleGAN Image Restoration
paper：https://arxiv.org/abs/2302.06733
[4]Raw Image Reconstruction with Learned Compact Metadata
paper：https://arxiv.org/abs/2302.12995
[5]Efficient and Explicit Modelling of Image Hierarchies for Image Restoration
paper：https://arxiv.org/abs/2303.00748
code：https://github.com/ofsoundof/GRL-Image-Restoration
[6]Imagic: Text-Based Real Image Editing with Diffusion Models
paper：https://arxiv.org/abs/2210.09276
project：https://imagic-editing.github.io/
[7]High-resolution image reconstruction with latent diffusion models from human brain activity
paper：https://www.biorxiv.org/content/10.1101/2022.11.18.517004v2
project：https://sites.google.com/view/stablediffusion-with-brain/
[8]Solving 3D Inverse Problems using Pre-trained 2D Diffusion Models
paper：https://arxiv.org/abs/2211.10655
图像去噪/去模糊/去雨去雾(Image Denoising
[1]Uncertainty-Aware Unsupervised Image Deblurring with Deep Residual Prior
paper：https://arxiv.org/abs/2210.05361
[2]Polarized Color Image Denoising using Pocoformer
paper：https://arxiv.org/abs/2207.00215
[3]Blur Interpolation Transformer for Real-World Motion from Blur
paper：https://arxiv.org/abs/2211.11423
code：https://github.com/zzh-tech/BiT
[4]Structured Kernel Estimation for Photon-Limited Deconvolution
paper：https://arxiv.org/abs/2303.03472
code：https://github.com/sanghviyashiitb/structured-kernel-cvpr23
图像编辑/图像修复(Image Edit/Inpainting
[1]LANIT: Language-Driven Image-to-Image Translation for Unlabeled Data
paper：https://arxiv.org/abs/2208.14889
code：https://github.com/KU-CVLAB/LANIT
图像质量评估(Image Quality Assessment
[1]CR-FIQA: Face Image Quality Assessment by Learning Sample Relative Classifiability
paper：https://arxiv.org/abs/2112.06592
[2]Quality-aware Pre-trained Models for Blind Image Quality Assessment
paper：https://arxiv.org/abs/2303.00521
图像配准(Image Registration
[1]Indescribable Multi-modal Spatial Evaluator
paper：https://arxiv.org/abs/2303.00369
code：https://github.com/Kid-Liet/IMSE/pulse
人脸(Face
人脸生成/合成/重建/编辑(Face Generation/Face Synthesis/Face Reconstruction/Face Editing
[1]A Hierarchical Representation Network for Accurate and Detailed Face Reconstruction from In-The-Wild Images
paper：https://arxiv.org/abs/2302.14434
[2]MetaPortrait: Identity-Preserving Talking Head Generation with Fast Personalized Adaptation(MetaPortrait：具有快速个性化适应的身份保持谈话头像生成
paper：https://arxiv.org/abs/2212.08062
code：https://github.com/Meta-Portrait/MetaPortrait
人脸伪造/反欺骗(Face Forgery/Face Anti-Spoofing
[1]Physical-World Optical Adversarial Attacks on 3D Face Recognition
paper：https://arxiv.org/abs/2205.13412
医学影像(Medical Imaging
[1]Deep Feature In-painting for Unsupervised Anomaly Detection in X-ray Images
paper：https://arxiv.org/pdf/2111.13495.pdf
code：https://github.com/tiangexiang/SQUID
[2]Label-Free Liver Tumor Segmentation
paper：https://arxiv.org/pdf/2210.14845.pdf
code：https://github.com/MrGiovanni/SyntheticTumors
图像生成/图像合成(Image Generation/Image Synthesis
[1]DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation
paper：https://arxiv.org/abs/2208.12242
code：https://github.com/PaddlePaddle/PaddleNLP/tree/develop/ppdiffusers/examples/dreambooth
[2]Progressive Open Space Expansion for Open-Set Model Attribution
paper：https://arxiv.org/abs/2303.06877
code：https://github.com/tianyunyoung/pose
[3]Person Image Synthesis via Denoising Diffusion Model
paper：https://arxiv.org/abs/2211.12500
[4]Solving 3D Inverse Problems using Pre-trained 2D Diffusion Models(使用预训练的 2D 扩散模型解决 3D 逆问题
paper：https://arxiv.org/abs/2211.10655
[5]Parallel Diffusion Models of Operator and Image for Blind Inverse Problems(盲反问题算子和图像的并行扩散模型
paper：https://arxiv.org/abs/2211.10656
场景重建/视图合成/新视角合成(Novel View Synthesis
[1]3D Video Loops from Asynchronous Input
paper：https://arxiv.org/abs/2303.05312
code：https://github.com/limacv/VideoLoop3D
[2]NeRFLiX: High-Quality Neural View Synthesis by Learning a Degradation-Driven Inter-viewpoint MiXer
paper：https://arxiv.org/abs/2303.06919
code：https://t.co/uNiTd9ujCv
[3]NeRF-Gaze: A Head-Eye Redirection Parametric Model for Gaze Estimation
paper：https://arxiv.org/abs/2212.14710
[4]Renderable Neural Radiance Map for Visual Navigation
paper：https://arxiv.org/abs/2303.00304
[5]Real-Time Neural Light Field on Mobile Devices
paper：https://arxiv.org/abs/2212.08057
project：https://snap-research.github.io/MobileR2L/
[6]Latent-NeRF for Shape-Guided Generation of 3D Shapes and Textures
paper：https://arxiv.org/abs/2211.07600
code：https://github.com/eladrich/latent-nerf
[7]NoPe-NeRF: Optimising Neural Radiance Field with No Pose Prior
paper：https://arxiv.org/abs/2212.07388
project：https://nope-nerf.active.vision/
多模态学习(Multi-Modal Learning
[1]Align and Attend: Multimodal Summarization with Dual Contrastive Losses
paper：https://arxiv.org/abs/2303.07284
code：https://boheumd.github.io/A2Summ/
[2]Towards All-in-one Pre-training via Maximizing Multi-modal Mutual Information(通过最大化多模态互信息实现一体化预训练
paper：https://arxiv.org/abs/2211.09807
code：https://github.com/OpenGVLab/M3I-Pretraining
[3]Uni-Perceiver v2: A Generalist Model for Large-Scale Vision and Vision-Language Tasks(Uni-Perceiver v2：用于大规模视觉和视觉语言任务的通才模型
paper：https://arxiv.org/abs/2211.09808
code：https://github.com/fundamentalvision/Uni-Perceiver
半监督学习/弱监督学习/无监督学习/自监督学习(Self-supervised Learning/Semi-supervised Learning)
[1]The Dialog Must Go On: Improving Visual Dialog via Generative Self-Training
paper：https://arxiv.org/abs/2205.12502
code：https://github.com/gicheonkang/gst-visdial
[2]Three Guidelines You Should Know for Universally Slimmable Self-Supervised Learning
paper：https://arxiv.org/abs/2303.06870
code：https://github.com/megvii-research/US3L-CVPR2023
[3]Mask3D: Pre-training 2D Vision Transformers by Learning Masked 3D Priors
paper：https://arxiv.org/abs/2302.14746
[4]Siamese Image Modeling for Self-Supervised Vision Representation Learning
paper：https://arxiv.org/abs/2206.01204
code：https://github.com/fundamentalvision/Siamese-Image-Modeling
[5]Cut and Learn for Unsupervised Object Detection and Instance Segmentation
paper：https://arxiv.org/abs/2301.11320
project：http://people.eecs.berkeley.edu
