CVPR2018で気になった論文の抜粋

wifiの電波強度から姿勢推定

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Through-Wall_Human_Pose_Estimation_Using_Radio_Signals

 RGBから3D姿勢推定

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Learning_to_Estimate_3D_Human_Pose_and_Shape_from_a_Single_Color_Image

#イベントの検出と分割

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/End-to-End_Dense_Video_Captioning_with_Masked_Transformer

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Ordinal_Depth_Supervision_for_3D_Human_Pose_Estimation

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Variational_Autoencoders_for_Deforming_3D_Mesh_Models

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/PoseFlow_A_Deep_Motion_Representation_for_Understanding_Human_Behaviors_in_Videos

繰り返しの動作を推定

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Real-World_Repetition_Estimation_by_Div_Grad_and_Curl

ゲームからのレンダリング手法

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Free_supervision_from_video_games

異常検知

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Real-world_Anomaly_Detection_in_Surveillance_Videos

距離画像推定

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Aperture_Supervision_for_Monocular_Depth_Estimation

3次元推定の比較実験

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Pixels_voxels_and_views_A_study_of_shape_representations_for_single_view_3D_object_shape_prediction

2D画像に移っている人物を色々な姿勢に変形！めっちゃ面白い！

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/A_Variational_U-Net_for_Conditional_Appearance_and_Shape_Generation

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Deformable_GANs_for_Pose-based_Human_Image_Generation

どっちの技能が優れているか！めっちゃ近い分野！！

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Whos_Better_Whos_Best_Pairwise_Deep_Ranking_for_Skill_Determination

動画から，アフォーダンス取得だけど，まだまだって感じ

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Demo2Vec_Reasoning_Object_Affordances_from_Online_Videos

2Dからの3次元姿勢推定

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Three_Dimension_Human_Pose_Estimation_in_the_Wild_by_Adversarial_Learning

長期のモーション推定

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Convolutional_Sequence_to_Sequence_Model_for_Human_Dynamics

動画から対象と同じアクションを探す

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/One-shot_Action_Localization_by_Learning_Sequence_Matching_Network

動画の行動のラベル・開始と終了を推定する

※ねじ締め，配線，といった単位で分割できるようになると面白いかも

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Rethinking_the_Faster_R-CNN_Architecture_for_Temporal_Action_Localization

動画から人物の3次元姿勢推定

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Video_Based_Reconstruction_of_3D_People_Models

動画から行動推定

https://cvpaperchallenge.github.io/CVPR2018_Survey/#/Weakly-Supervised_Action_Segmentation_with_Iterative_Soft_Boundary_Assignment