作为计算机视觉领域三大顶会之一,CVPR2021目前已公布了所有接收论文ID,一共有1663篇论文被接收,接收率为23.7%,虽然接受率相比去年有所上升,但竞争也是非常激烈。
CVPR2021 最全整理:论文分类汇总 / 代码 / 项目 / 论文解读(更新中):
https://bbs.cvmart.net/post/4267
此前我们对CVPR2020/2019/2018、ECCV2020、ICCV进行了分类汇总整理,所有的内容都汇总于社区 or Github:
https://bbs.cvmart.net/post/62
https://github.com/extreme-assistant/CVPR2021-Paper-Code-Interpretation
在本文中,我们对CVPR2021的最新论文进行了分类汇总,按研究方向整理。包含目标检测、图像分割、目标跟踪、医学影像、3D、模型压缩、图像处理、姿态估计、文本检测等多个方向,同时,我们将对优秀论文解读报道和技术直播,欢迎大家关注~
Instance Localization for Self-supervised Detection Pretraining
paper|code
Multiple Instance Active Learning for Object Detection(用于对象检测的多实例主动学习)
paper|code
Open-world object detection(开放世界中的目标检测)
code
Positive-Unlabeled Data Purification in the Wild for Object Detection(野外检测对象的阳性无标签数据提纯)
UP-DETR: Unsupervised Pre-training for Object Detection with Transformers
paper
解读:无监督预训练检测器
CanonPose: Self-supervised Monocular 3D Human Pose Estimation in the Wild(野外自监督的单眼3D人类姿态估计)
PCLs: Geometry-aware Neural Reconstruction of 3D Pose with Perspective Crop Layers(具有透视作物层的3D姿势的几何感知神经重建)
paper
Probabilistic Tracklet Scoring and Inpainting for Multiple Object Tracking(多目标跟踪的概率小波计分和修复)
paper
Rotation Equivariant Siamese Networks for Tracking(旋转等距连体网络进行跟踪)
paper
3D Graph Anatomy Geometry-Integrated Network for Pancreatic Mass Segmentation, Diagnosis, and Quantitative Patient Management(用于胰腺肿块分割,诊断和定量患者管理的3D图形解剖学几何集成网络)
Deep Lesion Tracker: Monitoring Lesions in 4D Longitudinal Imaging Studies(深部病变追踪器:在4D纵向成像研究中监控病变)
paper
Automatic Vertebra Localization and Identification in CT by Spine Rectification and Anatomically-constrained Optimization(通过脊柱矫正和解剖学约束优化在CT中自动进行椎骨定位和识别)
paper
AttentiveNAS: Improving Neural Architecture Search via Attentive(通过注意力改善神经架构搜索)
paper
ReNAS: Relativistic Evaluation of Neural Architecture Search(NAS predictor当中ranking loss的重要性)
paper
HourNAS: Extremely Fast Neural Architecture Search Through an Hourglass Lens(降低NAS的成本)
paper
Exploiting Spatial Dimensions of Latent in GAN for Real-time Image Editing(利用GAN中潜在的空间维度进行实时图像编辑)
Hijack-GAN: Unintended-Use of Pretrained, Black-Box GANs(Hijack-GAN:意外使用经过预训练的黑匣子GAN)
paper
Encoding in Style: a StyleGAN Encoder for Image-to-Image Translation(样式编码:用于图像到图像翻译的StyleGAN编码器)
paper|code|project
A 3D GAN for Improved Large-pose Facial Recognition(用于改善大姿势面部识别的3D GAN)
paper
Data-Free Knowledge Distillation For Image Super-Resolution(DAFL算法的SR版本)
AdderSR: Towards Energy Efficient Image Super-Resolution(将加法网路应用到图像超分辨率中)
paper|code
解读:华为开源加法神经网络
Manifold Regularized Dynamic Network Pruning(动态剪枝的过程中考虑样本复杂度与网络复杂度的约束)
Learning Student Networks in the Wild(一种不需要原始训练数据的模型压缩和加速技术)
paper|code
解读:华为诺亚方舟实验室提出无需数据网络压缩技术
Multiresolution Knowledge Distillation for Anomaly Detection(用于异常检测的多分辨率知识蒸馏)
paper
Distilling Object Detectors via Decoupled Features(前景背景分离的蒸馏技术)
Rethinking Channel Dimensions for Efficient Model Design(重新考虑通道尺寸以进行有效的模型设计)
paper|code
Inverting the Inherence of Convolution for Visual Recognition(颠倒卷积的固有性以进行视觉识别)
RepVGG: Making VGG-style ConvNets Great Again
paper|code
解读:RepVGG:极简架构,SOTA性能,让VGG式模型再次伟大
Transformer Interpretability Beyond Attention Visualization(注意力可视化之外的Transformer可解释性)
paper|code
UP-DETR: Unsupervised Pre-training for Object Detection with Transformers
paper
解读:无监督预训练检测器
Pre-Trained Image Processing Transformer(底层视觉预训练模型)
paper
Sequential Graph Convolutional Network for Active Learning(主动学习的顺序图卷积网络)
paper
Meta Batch-Instance Normalization for Generalizable Person Re-Identification(通用批处理人员重新标识的元批实例规范化)
paper
Representative Batch Normalization with Feature Calibration(具有特征校准功能的代表性批量归一化)
Multiple Instance Active Learning for Object Detection(用于对象检测的多实例主动学习)
paper|code
Sequential Graph Convolutional Network for Active Learning(主动学习的顺序图卷积网络)
paper
Few-shot Open-set Recognition by Transformation Consistency(转换一致性很少的开放集识别)
Exploring Complementary Strengths of Invariant and Equivariant Representations for Few-Shot Learning(探索少量学习的不变表示形式和等变表示形式的互补强度)
Rainbow Memory: Continual Learning with a Memory of Diverse Samples(不断学习与多样本的记忆)
Learning the Superpixel in a Non-iterative and Lifelong Manner(以非迭代和终身的方式学习超像素)
Diversifying Sample Generation for Data-Free Quantization(多样化的样本生成,实现无数据量化)
paper
Domain Generalization via Inference-time Label-Preserving Target Projections(通过保留推理时间的目标投影进行域泛化)
paper
DeRF: Decomposed Radiance Fields(分解的辐射场)
project
Vab-AL: Incorporating Class Imbalance and Difficulty with Variational Bayes for Active Learning(将类不平衡和复杂性与变式贝叶斯结合起来进行主动学习)
paper
Densely connected multidilated convolutional networks for dense prediction tasks(密集连接的多重卷积网络,用于密集的预测任务)
paper
VirTex: Learning Visual Representations from Textual Annotations(从文本注释中学习视觉表示)
paper|code
Improving Unsupervised Image Clustering With Robust Learning(通过鲁棒学习改善无监督图像聚类)
paper|code
Weakly-supervised Grounded Visual Question Answering using Capsules(使用胶囊进行弱监督的地面视觉问答)
FLAVR: Flow-Agnostic Video Representations for Fast Frame Interpolation(FLAVR:用于快速帧插值的与流无关的视频表示)
paper|code|project
Probabilistic Embeddings for Cross-Modal Retrieval(跨模态检索的概率嵌入)
paper
Self-supervised Simultaneous Multi-Step Prediction of Road Dynamics and Cost Map(道路动力学和成本图的自监督式多步同时预测)
IIRC: Incremental Implicitly-Refined Classification(增量式隐式定义的分类)
paper|project
Fair Attribute Classification through Latent Space De-biasing(通过潜在空间去偏的公平属性分类)
paper|code|project
Information-Theoretic Segmentation by Inpainting Error Maximization(修复误差最大化的信息理论分割)
paper
UC2: Universal Cross-lingual Cross-modal Vision-and-Language Pretraining(UC2:通用跨语言跨模态视觉和语言预培训)
Less is More: CLIPBERT for Video-and-Language Learning via Sparse Sampling(T通过稀疏采样进行视频和语言学习)
paper|code
D-NeRF: Neural Radiance Fields for Dynamic Scenes(D-NeRF:动态场景的神经辐射场)
paper|project
Weakly Supervised Learning of Rigid 3D Scene Flow(刚性3D场景流的弱监督学习)
paper|code|project