The Experts below are selected from a list of 12351 Experts worldwide ranked by ideXlab platform
Adam Kortylewski - One of the best experts on this subject based on the ideXlab platform.
-
Compositional Convolutional Neural Networks: A Robust and Interpretable Model for Object Recognition Under Occlusion
International Journal of Computer Vision, 2020Co-Authors: Adam Kortylewski, Qing Liu, Angtian Wang, Yihong Sun, Alan YuilleAbstract:Computer vision systems in real-world applications need to be robust to Partial Occlusion while also being explainable. In this work, we show that black-box deep convolutional neural networks (DCNNs) have only limited robustness to Partial Occlusion. We overcome these limitations by unifying DCNNs with part-based models into Compositional Convolutional Neural Networks (CompositionalNets)—an interpretable deep architecture with innate robustness to Partial Occlusion. Specifically, we propose to replace the fully connected classification head of DCNNs with a differentiable compositional model that can be trained end-to-end. The structure of the compositional model enables CompositionalNets to decompose images into objects and context, as well as to further decompose object representations in terms of individual parts and the objects’ pose. The generative nature of our compositional model enables it to localize occluders and to recognize objects based on their non-occluded parts. We conduct extensive experiments in terms of image classification and object detection on images of artificially occluded objects from the PASCAL3D+ and ImageNet dataset, and real images of Partially occluded vehicles from the MS-COCO dataset. Our experiments show that CompositionalNets made from several popular DCNN backbones (VGG-16, ResNet50, ResNext) improve by a large margin over their non-compositional counterparts at classifying and detecting Partially occluded objects. Furthermore, they can localize occluders accurately despite being trained with class-level supervision only. Finally, we demonstrate that CompositionalNets provide human interpretable predictions as their individual components can be understood as detecting parts and estimating an objects’ viewpoint.
-
compositional convolutional neural networks a deep architecture with innate robustness to Partial Occlusion
Computer Vision and Pattern Recognition, 2020Co-Authors: Adam Kortylewski, Qing Liu, Alan L YuilleAbstract:Recent work has shown that deep convolutional neural networks (DCNNs) do not generalize well under Partial Occlusion. Inspired by the success of compositional models at classifying Partially occluded objects, we propose to integrate compositional models and DCNNs into a unified deep model with innate robustness to Partial Occlusion. We term this architecture Compositional Convolutional Neural Network. In particular, we propose to replace the fully connected classification head of a DCNN with a differentiable compositional model. The generative nature of the compositional model enables it to localize occluders and subsequently focus on the non-occluded parts of the object. We conduct classification experiments on artificially occluded images as well as real images of Partially occluded objects from the MS-COCO dataset. The results show that DCNNs do not classify occluded objects robustly, even when trained with data that is strongly augmented with Partial Occlusions. Our proposed model outperforms standard DCNNs by a large margin at classifying Partially occluded objects, even when it has not been exposed to occluded objects during training. Additional experiments demonstrate that CompositionalNets can also localize the occluders accurately, despite being trained with class labels only. The code and data used in this work are publicly available.
-
compositional convolutional neural networks a deep architecture with innate robustness to Partial Occlusion
arXiv: Computer Vision and Pattern Recognition, 2020Co-Authors: Adam Kortylewski, Qing Liu, Alan L YuilleAbstract:Recent findings show that deep convolutional neural networks (DCNNs) do not generalize well under Partial Occlusion. Inspired by the success of compositional models at classifying Partially occluded objects, we propose to integrate compositional models and DCNNs into a unified deep model with innate robustness to Partial Occlusion. We term this architecture Compositional Convolutional Neural Network. In particular, we propose to replace the fully connected classification head of a DCNN with a differentiable compositional model. The generative nature of the compositional model enables it to localize occluders and subsequently focus on the non-occluded parts of the object. We conduct classification experiments on artificially occluded images as well as real images of Partially occluded objects from the MS-COCO dataset. The results show that DCNNs do not classify occluded objects robustly, even when trained with data that is strongly augmented with Partial Occlusions. Our proposed model outperforms standard DCNNs by a large margin at classifying Partially occluded objects, even when it has not been exposed to occluded objects during training. Additional experiments demonstrate that CompositionalNets can also localize the occluders accurately, despite being trained with class labels only. The code used in this work is publicly available.
-
combining compositional models and deep networks for robust object classification under Occlusion
Workshop on Applications of Computer Vision, 2020Co-Authors: Adam Kortylewski, Qing Liu, Huiyu Wang, Zhishuai Zhang, Alan L YuilleAbstract:Deep convolutional neural networks (DCNNs) are powerful models that yield impressive results at object classification. However, recent work has shown that they do not generalize well to Partially occluded objects and to mask attacks. In contrast to DCNNs, compositional models are robust to Partial Occlusion, however, they are not as discriminative as deep models. In this work, we combine DC-NNs and compositional object models to retain the best of both approaches: a discriminative model that is robust to Partial Occlusion and mask attacks. Our model is learned in two steps. First, a standard DCNN is trained for image classification. Subsequently, we cluster the DCNN features into dictionaries. We show that the dictionary components resemble object part detectors and learn the spatial distribution of parts for each object class. We propose mixtures of compositional models to account for large changes in the spatial activation patterns (e.g. due to changes in the 3D pose of an object). At runtime, an image is first classified by the DCNN in a feedforward manner. The prediction uncertainty is used to detect Partially occluded objects, which in turn are classified by the compositional model. Our experimental results demonstrate that combining compositional models and DCNNs resolves a fundamental problem of current deep learning approaches to computer vision: The combined model recognizes occluded objects, even when it has not been exposed to occluded objects during training, while at the same time maintaining high discriminative performance for non-occluded objects.
Alan L Yuille - One of the best experts on this subject based on the ideXlab platform.
-
compositional convolutional neural networks a deep architecture with innate robustness to Partial Occlusion
Computer Vision and Pattern Recognition, 2020Co-Authors: Adam Kortylewski, Qing Liu, Alan L YuilleAbstract:Recent work has shown that deep convolutional neural networks (DCNNs) do not generalize well under Partial Occlusion. Inspired by the success of compositional models at classifying Partially occluded objects, we propose to integrate compositional models and DCNNs into a unified deep model with innate robustness to Partial Occlusion. We term this architecture Compositional Convolutional Neural Network. In particular, we propose to replace the fully connected classification head of a DCNN with a differentiable compositional model. The generative nature of the compositional model enables it to localize occluders and subsequently focus on the non-occluded parts of the object. We conduct classification experiments on artificially occluded images as well as real images of Partially occluded objects from the MS-COCO dataset. The results show that DCNNs do not classify occluded objects robustly, even when trained with data that is strongly augmented with Partial Occlusions. Our proposed model outperforms standard DCNNs by a large margin at classifying Partially occluded objects, even when it has not been exposed to occluded objects during training. Additional experiments demonstrate that CompositionalNets can also localize the occluders accurately, despite being trained with class labels only. The code and data used in this work are publicly available.
-
compositional convolutional neural networks a deep architecture with innate robustness to Partial Occlusion
arXiv: Computer Vision and Pattern Recognition, 2020Co-Authors: Adam Kortylewski, Qing Liu, Alan L YuilleAbstract:Recent findings show that deep convolutional neural networks (DCNNs) do not generalize well under Partial Occlusion. Inspired by the success of compositional models at classifying Partially occluded objects, we propose to integrate compositional models and DCNNs into a unified deep model with innate robustness to Partial Occlusion. We term this architecture Compositional Convolutional Neural Network. In particular, we propose to replace the fully connected classification head of a DCNN with a differentiable compositional model. The generative nature of the compositional model enables it to localize occluders and subsequently focus on the non-occluded parts of the object. We conduct classification experiments on artificially occluded images as well as real images of Partially occluded objects from the MS-COCO dataset. The results show that DCNNs do not classify occluded objects robustly, even when trained with data that is strongly augmented with Partial Occlusions. Our proposed model outperforms standard DCNNs by a large margin at classifying Partially occluded objects, even when it has not been exposed to occluded objects during training. Additional experiments demonstrate that CompositionalNets can also localize the occluders accurately, despite being trained with class labels only. The code used in this work is publicly available.
-
combining compositional models and deep networks for robust object classification under Occlusion
Workshop on Applications of Computer Vision, 2020Co-Authors: Adam Kortylewski, Qing Liu, Huiyu Wang, Zhishuai Zhang, Alan L YuilleAbstract:Deep convolutional neural networks (DCNNs) are powerful models that yield impressive results at object classification. However, recent work has shown that they do not generalize well to Partially occluded objects and to mask attacks. In contrast to DCNNs, compositional models are robust to Partial Occlusion, however, they are not as discriminative as deep models. In this work, we combine DC-NNs and compositional object models to retain the best of both approaches: a discriminative model that is robust to Partial Occlusion and mask attacks. Our model is learned in two steps. First, a standard DCNN is trained for image classification. Subsequently, we cluster the DCNN features into dictionaries. We show that the dictionary components resemble object part detectors and learn the spatial distribution of parts for each object class. We propose mixtures of compositional models to account for large changes in the spatial activation patterns (e.g. due to changes in the 3D pose of an object). At runtime, an image is first classified by the DCNN in a feedforward manner. The prediction uncertainty is used to detect Partially occluded objects, which in turn are classified by the compositional model. Our experimental results demonstrate that combining compositional models and DCNNs resolves a fundamental problem of current deep learning approaches to computer vision: The combined model recognizes occluded objects, even when it has not been exposed to occluded objects during training, while at the same time maintaining high discriminative performance for non-occluded objects.
Qing Liu - One of the best experts on this subject based on the ideXlab platform.
-
Compositional Convolutional Neural Networks: A Robust and Interpretable Model for Object Recognition Under Occlusion
International Journal of Computer Vision, 2020Co-Authors: Adam Kortylewski, Qing Liu, Angtian Wang, Yihong Sun, Alan YuilleAbstract:Computer vision systems in real-world applications need to be robust to Partial Occlusion while also being explainable. In this work, we show that black-box deep convolutional neural networks (DCNNs) have only limited robustness to Partial Occlusion. We overcome these limitations by unifying DCNNs with part-based models into Compositional Convolutional Neural Networks (CompositionalNets)—an interpretable deep architecture with innate robustness to Partial Occlusion. Specifically, we propose to replace the fully connected classification head of DCNNs with a differentiable compositional model that can be trained end-to-end. The structure of the compositional model enables CompositionalNets to decompose images into objects and context, as well as to further decompose object representations in terms of individual parts and the objects’ pose. The generative nature of our compositional model enables it to localize occluders and to recognize objects based on their non-occluded parts. We conduct extensive experiments in terms of image classification and object detection on images of artificially occluded objects from the PASCAL3D+ and ImageNet dataset, and real images of Partially occluded vehicles from the MS-COCO dataset. Our experiments show that CompositionalNets made from several popular DCNN backbones (VGG-16, ResNet50, ResNext) improve by a large margin over their non-compositional counterparts at classifying and detecting Partially occluded objects. Furthermore, they can localize occluders accurately despite being trained with class-level supervision only. Finally, we demonstrate that CompositionalNets provide human interpretable predictions as their individual components can be understood as detecting parts and estimating an objects’ viewpoint.
-
compositional convolutional neural networks a deep architecture with innate robustness to Partial Occlusion
Computer Vision and Pattern Recognition, 2020Co-Authors: Adam Kortylewski, Qing Liu, Alan L YuilleAbstract:Recent work has shown that deep convolutional neural networks (DCNNs) do not generalize well under Partial Occlusion. Inspired by the success of compositional models at classifying Partially occluded objects, we propose to integrate compositional models and DCNNs into a unified deep model with innate robustness to Partial Occlusion. We term this architecture Compositional Convolutional Neural Network. In particular, we propose to replace the fully connected classification head of a DCNN with a differentiable compositional model. The generative nature of the compositional model enables it to localize occluders and subsequently focus on the non-occluded parts of the object. We conduct classification experiments on artificially occluded images as well as real images of Partially occluded objects from the MS-COCO dataset. The results show that DCNNs do not classify occluded objects robustly, even when trained with data that is strongly augmented with Partial Occlusions. Our proposed model outperforms standard DCNNs by a large margin at classifying Partially occluded objects, even when it has not been exposed to occluded objects during training. Additional experiments demonstrate that CompositionalNets can also localize the occluders accurately, despite being trained with class labels only. The code and data used in this work are publicly available.
-
compositional convolutional neural networks a deep architecture with innate robustness to Partial Occlusion
arXiv: Computer Vision and Pattern Recognition, 2020Co-Authors: Adam Kortylewski, Qing Liu, Alan L YuilleAbstract:Recent findings show that deep convolutional neural networks (DCNNs) do not generalize well under Partial Occlusion. Inspired by the success of compositional models at classifying Partially occluded objects, we propose to integrate compositional models and DCNNs into a unified deep model with innate robustness to Partial Occlusion. We term this architecture Compositional Convolutional Neural Network. In particular, we propose to replace the fully connected classification head of a DCNN with a differentiable compositional model. The generative nature of the compositional model enables it to localize occluders and subsequently focus on the non-occluded parts of the object. We conduct classification experiments on artificially occluded images as well as real images of Partially occluded objects from the MS-COCO dataset. The results show that DCNNs do not classify occluded objects robustly, even when trained with data that is strongly augmented with Partial Occlusions. Our proposed model outperforms standard DCNNs by a large margin at classifying Partially occluded objects, even when it has not been exposed to occluded objects during training. Additional experiments demonstrate that CompositionalNets can also localize the occluders accurately, despite being trained with class labels only. The code used in this work is publicly available.
-
combining compositional models and deep networks for robust object classification under Occlusion
Workshop on Applications of Computer Vision, 2020Co-Authors: Adam Kortylewski, Qing Liu, Huiyu Wang, Zhishuai Zhang, Alan L YuilleAbstract:Deep convolutional neural networks (DCNNs) are powerful models that yield impressive results at object classification. However, recent work has shown that they do not generalize well to Partially occluded objects and to mask attacks. In contrast to DCNNs, compositional models are robust to Partial Occlusion, however, they are not as discriminative as deep models. In this work, we combine DC-NNs and compositional object models to retain the best of both approaches: a discriminative model that is robust to Partial Occlusion and mask attacks. Our model is learned in two steps. First, a standard DCNN is trained for image classification. Subsequently, we cluster the DCNN features into dictionaries. We show that the dictionary components resemble object part detectors and learn the spatial distribution of parts for each object class. We propose mixtures of compositional models to account for large changes in the spatial activation patterns (e.g. due to changes in the 3D pose of an object). At runtime, an image is first classified by the DCNN in a feedforward manner. The prediction uncertainty is used to detect Partially occluded objects, which in turn are classified by the compositional model. Our experimental results demonstrate that combining compositional models and DCNNs resolves a fundamental problem of current deep learning approaches to computer vision: The combined model recognizes occluded objects, even when it has not been exposed to occluded objects during training, while at the same time maintaining high discriminative performance for non-occluded objects.
Alan Yuille - One of the best experts on this subject based on the ideXlab platform.
-
Compositional Convolutional Neural Networks: A Robust and Interpretable Model for Object Recognition Under Occlusion
International Journal of Computer Vision, 2020Co-Authors: Adam Kortylewski, Qing Liu, Angtian Wang, Yihong Sun, Alan YuilleAbstract:Computer vision systems in real-world applications need to be robust to Partial Occlusion while also being explainable. In this work, we show that black-box deep convolutional neural networks (DCNNs) have only limited robustness to Partial Occlusion. We overcome these limitations by unifying DCNNs with part-based models into Compositional Convolutional Neural Networks (CompositionalNets)—an interpretable deep architecture with innate robustness to Partial Occlusion. Specifically, we propose to replace the fully connected classification head of DCNNs with a differentiable compositional model that can be trained end-to-end. The structure of the compositional model enables CompositionalNets to decompose images into objects and context, as well as to further decompose object representations in terms of individual parts and the objects’ pose. The generative nature of our compositional model enables it to localize occluders and to recognize objects based on their non-occluded parts. We conduct extensive experiments in terms of image classification and object detection on images of artificially occluded objects from the PASCAL3D+ and ImageNet dataset, and real images of Partially occluded vehicles from the MS-COCO dataset. Our experiments show that CompositionalNets made from several popular DCNN backbones (VGG-16, ResNet50, ResNext) improve by a large margin over their non-compositional counterparts at classifying and detecting Partially occluded objects. Furthermore, they can localize occluders accurately despite being trained with class-level supervision only. Finally, we demonstrate that CompositionalNets provide human interpretable predictions as their individual components can be understood as detecting parts and estimating an objects’ viewpoint.
Henri Hepner - One of the best experts on this subject based on the ideXlab platform.
-
aneurysm clipping after endovascular treatment with coils a report of eight patients
Neurosurgery, 1996Co-Authors: Thierry Civit, Jean Auque, Serge Bracard, Jean-claude Marchal, L Picard, Henri HepnerAbstract:Between January 1990 and December 1994, patients with subarachnoid hemorrhage related to ruptured aneurysms who were referred to our institution were treated by neurosurgical and neuroradiological teams. In each patient, the respective indications for neurosurgical or endovascular treatment were discussed, taking into consideration patients' age and the morphological and topographical aneurysm features. We report eight cases of patients with subarachnoid hemorrhage who underwent operations after primary endovascular procedures (Hunt and Hess scores III, IV, and V). The indications for surgical treatment were as follows. First, deliberate Partial Occlusion of the aneurysm (two aneurysms of the internal carotid artery and one aneurysm of the anterior communicating artery) was performed to obtain only Partial clotting of the aneurysm sac by free coils. However, this procedure was discontinued in favor of the use of Guglielmi detachable coils. The second indication was Partial Occlusion after an endovascular procedure (two aneurysms of the middle cerebral artery and one internal carotid artery aneurysm). The third indication was re-expansion of the aneurysm 1 year after the endovascular treatment (one middle cerebral artery aneurysm). The final indication was secondary rupture of the aneurysm sac and false aneurysm around the migrating coil (one aneurysm of the pericallosal artery). During surgery, the aneurysm sac appeared translucent. The coils bulged out and stretched the aneurysm sac. One ruptured the membrane leading to a subarachnoid hemorrhage during the endovascular procedure. No hemorrhage occurred during the surgical clipping. Aneurysm obliteration was easily performed, especially when the packing was Partial, but was very difficult when the complete aneurysm closure led to a stenosis of the parent vessel. A giant sylvian aneurysm rest, visible only with angiography, was left untreated. This series illustrates an original experience, which led us to conclude that aneurysm surgery with coils in place is not as difficult as is often thought.