每月见闻202110

发表于 |

ICCV2021

不久前计算机视觉三大顶会之一的ICCV2021接收结果已经公布，本次ICCV共计6236篇有效提交论文，其中有1617篇被接收，接收率为25.9%

论文共分为检测、分割、估计、跟踪、视觉定位、底层图像处理、图像视频检索、三位视觉等多个方向，极市团队-ICCV论文代码整理该Github项目整理了所有的相关论文和代码，主要有以下方向

  • 2D Object Detection(2D目标检测)

  • 3D Object Detection(3D目标检测)

  • Saliency Object Detection(显著性目标检测)

  • Camouflaged Object Detection(伪装目标检测)

  • Anomally Detection in Image(图像异常检测/表面缺陷检测)

  • Edge Detection(边缘检测)

  • Image Segmentation(图像分割)

  • Instance Segmentation(示例分割)

  • Semantic Segmentation(语义分割)

  • Video Object Segmentation(视频目标分割)

  • Referring Image Segmentation(参考图像分割)

  • Dense Prediction(密集预测)

  • Facial Recognition/Detection(人脸识别/检测)

  • Face Generation/Face Synthesis/Face Reconstruction/Face Editing(人脸生成/合成/重建/编辑)

  • Face Forgery/Face Anti-Spoofing(人脸伪造/反欺骗)

  • 3D Vision(三维视觉)

  • Point Cloud(点云)

  • 3D Reconstruction(三维重建)

  • Neural Network Structure Design & Optimization(神经网络设计与优化)

  • Transformer

  • NAS(神经网络架构搜索)

  • Loss Function(损失函数)

  • Visualization/Interpretability(可视化/可解释性)

  • Model Training/Generalization(模型训练/泛化)

  • Noisy Label(噪声标签)

  • Long-Tailed Distribution(长尾分布)

  • Out of Distribution Detection(分布外样本检测)

  • Knowledge Distillation(知识蒸馏)

  • Pruning(剪枝)

  • Quantization(量化)

  • Image Generation/Image Synthesis(图像生成/合成)

  • View Synthesis(视图合成)

  • GAN/Generative/Adversarial(GAN/生成式/对抗式)

  • Image Processing(图像处理)

  • Super Resolution(超分辨率)

  • Image Denoising(图像去噪/去模糊/去雨去雾)

  • Image Edit/Image Inpainting(图像编辑/修复)

  • Style Transfer(风格迁移)

  • Image Quality Assessment(图像质量评估)

  • Human Pose Estimation(姿态估计)

  • Depth Estimation(深度估计)

  • Image&Video Retrieval/Video Understanding(图像&视频检索/理解)

  • Action/Activity Recognition(行为识别/动作识别)

  • Re-Identification/Detection(行人重识别/检测)

  • Image/Video Caption(图像/视频字幕)

  • Visual Localization(视觉定位)

  • Image Matching(图像匹配)

  • 3D Vision(三维视觉)

  • Object Tracking(目标跟踪)

  • Medical Imaging(医学影像)

  • Text Detection/Recognition(文本检测/识别)

  • Remote Sensing Image(遥感图像)

  • Scene Graph Generation(场景图生成)

  • Scene Graph Prediction(场景图预测)

  • Data Augmentation(数据增广)

  • Anomaly Detection(异常检测)

  • Representation Learning(表征学习)

  • Image Clustering(图像聚类)

  • Few-shot/Zero-shot Learning(小样本学习/零样本学习)

  • Continual Learning/Life-long Learning(持续学习)

  • Transfer Learning/Domain Adaptation(迁移学习/自适应)

  • Metric Learning(度量学习)

  • Incremental Learning(增量学习)

  • Contrastive Learning(对比学习)

  • Visual Reasoning/VQA(推理学习/视觉问答)