Deformable Convolutional Networks
ICCV17 | 133 | Deformable Convolutional Networks
Jifeng Dai (Microsoft Research), Haozhi Qi (microsoft.com; ust.hk), Yuwen Xiong (microsoft.com; zju.edu.cn), Yi Li (microsoft.com; tsinghua.edu.cn), Guodong Zhang (microsoft.com; zju.edu.cn), Han Hu (Microsoft Research Asia), Yichen Wei ()
Convolutional neural networks (CNNs) are inherently limited to model geometric transformations due to the fixed geometric structures in its building modules. In this work, we introduce two new modules to enhance the transformation modeling capacity of CNNs, namely, deformable convolution and deformable RoI pooling. Both are based on the idea of augmenting the spatial sampling locations in the modules with additional offsets and learning the offsets from target tasks, without additional supervision. The new modules can readily replace their plain counterparts in existing CNNs and can be easily trained end-to-end by standard back-propagation, giving rise to deformable convolutional networks. Extensive experiments validate the effectiveness of our approach on sophisticated vision tasks of object detection and semantic segmentation. The code would be released.
Видео Deformable Convolutional Networks канала ComputerVisionFoundation Videos
Jifeng Dai (Microsoft Research), Haozhi Qi (microsoft.com; ust.hk), Yuwen Xiong (microsoft.com; zju.edu.cn), Yi Li (microsoft.com; tsinghua.edu.cn), Guodong Zhang (microsoft.com; zju.edu.cn), Han Hu (Microsoft Research Asia), Yichen Wei ()
Convolutional neural networks (CNNs) are inherently limited to model geometric transformations due to the fixed geometric structures in its building modules. In this work, we introduce two new modules to enhance the transformation modeling capacity of CNNs, namely, deformable convolution and deformable RoI pooling. Both are based on the idea of augmenting the spatial sampling locations in the modules with additional offsets and learning the offsets from target tasks, without additional supervision. The new modules can readily replace their plain counterparts in existing CNNs and can be easily trained end-to-end by standard back-propagation, giving rise to deformable convolutional networks. Extensive experiments validate the effectiveness of our approach on sophisticated vision tasks of object detection and semantic segmentation. The code would be released.
Видео Deformable Convolutional Networks канала ComputerVisionFoundation Videos
Показать
Комментарии отсутствуют
Информация о видео
17 ноября 2017 г. 3:29:42
00:11:47
Другие видео канала
Lecture 11 | Detection and SegmentationConvolutional Neural Networks (CNNs) explainedShaping the Latent Space for Data Interpolation - Vision Day 2018Event-based Asynchronous Sparse Convolutional NetworksKPConv MethodDeep Learning for 3D face reconstructionWhat is convolution? This is the easiest way to understandC 7.3 | Spatial Pyramid Pooling | SPPNet Classification | Fast RCNN | Machine learning | EvODNCapsule Networks (CapsNets) – TutorialObject Recognition With Deformable ModelsMask Region based Convolution Neural Networks - EXPLAINED!Torch Points3D — A unifying framework for deep learning on point clouds, Thomas ChatonDilated ConvolutionWACV18: A Rotationally-Invariant Convolution Module by Feature Map Back-Rotation6.5: TensorFlow.js: Layers API Part 1 - Intelligence and Learning3D Deep Learning TutorialPR-002: Deformable Convolutional Networks (2017)Deep Reinforcement Learning for Walking Robots - MATLAB and Simulink Robotics ArenaC4W1L05 Strided Convolutions