CVPR2022論文速遞（2022.4.8）！共18篇
語義分割/Segmentation - 3 篇
Pin the Memory: Learning to Generalize Semantic Segmentation
標題：針內存：學習概括語義細分
論文/Paper: http://arxiv.org/pdf/2204.03609
代碼/Code: None
Coarse-to-Fine Feature Mining for Video Semantic Segmentation
標題：用于視頻語義分割的粗對細特征挖掘
論文/Paper: http://arxiv.org/pdf/2204.03330
代碼/Code: https://github.com/guoleisun/vss-cffm
L2G: A Simple Local-to-Global Knowledge Transfer Framework for Weakly Supervised Semantic Segmentation
標題：L2G：一個簡單的本地對全球知識轉移框架，用于弱監(jiān)督語義分割
論文/Paper: http://arxiv.org/pdf/2204.03206
代碼/Code: https://github.com/PengtaoJiang/L2G.
GAN - 1 篇
Unsupervised Image-to-Image Translation with Generative Prior
標題：與生成的未經(jīng)監(jiān)督的圖像到圖像轉換
論文/Paper: http://arxiv.org/pdf/2204.03641
代碼/Code: https://github.com/williamyang1991/gp-unit
Transformers - - 1 篇
PSTR: End-to-End One-Step Person Search With Transformers
標題：PSTR：結束到最后的一步人與變壓器搜索
論文/Paper: http://arxiv.org/pdf/2204.03340
代碼/Code: https://github.com/jialecao001/pstr
對比學習/Contrastive Learning - 1 篇
Unified Contrastive Learning in Image-Text-Label Space
標題：在圖像文本標簽空間中的統(tǒng)一對比學習
論文/Paper: http://arxiv.org/pdf/2204.03610
代碼/Code: https://github.com/microsoft/unicl
視頻插幀/Frame Interpolation - 1 篇
Many-to-many Splatting for Efficient Video Frame Interpolation
標題：有效的視頻幀插值多對多分裂
論文/Paper: http://arxiv.org/pdf/2204.03513
代碼/Code: https://github.com/feinanshan/m2m_vfi
其他/Other - 11 篇
Total Variation Optimization Layers for Computer Vision
標題：計算機視覺的總變化優(yōu)化層
論文/Paper: http://arxiv.org/pdf/2204.03643
代碼/Code: https://github.com/raymondyeh07/tv_layers_for_cv
Pre-train, Self-train, Distill: A simple recipe for Supersizing 3D Reconstruction
標題：火車前，自動列車，蒸餾：一個簡單的配方，用于超出3D重建
論文/Paper: http://arxiv.org/pdf/2204.03642
代碼/Code: None
Class-Incremental Learning with Strong Pre-trained Models
標題：Class-Incremental學習與強大的預先訓練模型
論文/Paper: http://arxiv.org/pdf/2204.03634
代碼/Code: None
AutoRF: Learning 3D Object Radiance Fields from Single View Observations
標題：autorf：從單視圖觀察中學習3D對象輻射字段
論文/Paper: http://arxiv.org/pdf/2204.03593
代碼/Code: None
Deep Visual Geo-localization Benchmark
標題：深度視覺地理定位基準
論文/Paper: http://arxiv.org/pdf/2204.03444
代碼/Code: None
Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality
標題：Winoground：Visio-linguisticsitality的探測視覺和語言模型
論文/Paper: http://arxiv.org/pdf/2204.03162
代碼/Code: None
UIGR: Unified Interactive Garment Retrieval
標題：UIGR：統(tǒng)一互動服裝檢索
論文/Paper: http://arxiv.org/pdf/2204.03111
代碼/Code: https://github.com/brandonhanx/compfashion
AUV-Net: Learning Aligned UV Maps for Texture Transfer and Synthesis
標題：AUV-NET：學習對齊的UV地圖，用于紋理轉移和合成
論文/Paper: http://arxiv.org/pdf/2204.03105
代碼/Code: None
Hierarchical Self-supervised Representation Learning for Movie Understanding
標題：電影理解的分層自我監(jiān)督的代表學習
論文/Paper: http://arxiv.org/pdf/2204.03101
代碼/Code: None
Learning from Untrimmed Videos: Self-Supervised Video Representation Learning with Hierarchical Consistency
標題：從未經(jīng)監(jiān)控的視頻中學習：自我監(jiān)督視頻表示學習，具有分層一致性
論文/Paper: http://arxiv.org/pdf/2204.03017
代碼/Code: None
Multi-Scale Memory-Based Video Deblurring
標題：基于多尺度內存的視頻去紋理
論文/Paper: http://arxiv.org/pdf/2204.02977
代碼/Code: https://github.com/jibo27/memdeblur