다중혈관 관상동맥 환자에서 y-문합을 이용하여 양쪽 내흉동맥만을 사용한 우회술의 조기 성적

In this work, we revisit atrous convolution, a powerful tool to explicitly adjust filter's field-of-view as well as control the resolution of feature responses computed by Deep Convolutional Neural Networks, in the application of semantic image segmentation. To handle the problem of segmenting objects at multiple scales, we design modules which employ atrous convolution in cascade or in parallel to capture multi-scale context by adopting multiple atrous rates. Furthermore, we propose to augment our previously proposed Atrous Spatial Pyramid Pooling module, which probes convolutional features at multiple scales, with image-level features encoding global context and further boost performance. We also elaborate on implementation details and share our experience on training our system. The proposed `DeepLabv3' system significantly improves over our previous DeepLab versions without DenseCRF post-processing and attains comparable performance with other state-of-art models on the PASCAL VOC 2012 semantic image segmentation benchmark.

Rethinking Atrous Convolution for Semantic Image Segmentation

“Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks”の学習報告

The PASCAL Visual Object Classes Challenge

Semantic segmentation requires both rich spatial information and sizeable receptive field. However, modern approaches usually compromise spatial resolution to achieve real-time inference speed, which leads to poor performance. In this paper, we address this dilemma with a novel Bilateral Segmentation Network (BiSeNet). We first design a Spatial Path with a small stride to preserve the spatial information and generate high-resolution features. Meanwhile, a Context Path with a fast downsampling strategy is employed to obtain sufficient receptive field. On top of the two paths, we introduce a new Feature Fusion Module to combine features efficiently. The proposed architecture makes a right balance between the speed and segmentation performance on Cityscapes, CamVid, and COCO-Stuff datasets. Specifically, for a 2048 \(\times \) 1024 input, we achieve 68.4% Mean IOU on the Cityscapes test dataset with speed of 105 FPS on one NVIDIA Titan XP card, which is significantly faster than the existing methods with comparable performance.

/pdf/bisenet-bilateral-segmentation-network-for-real-time-2pc7rh0tw1.pdf

BiSeNet: Bilateral Segmentation Network for Real-time Semantic Segmentation

We participated in the high-level feature extraction task in TRECVID 2007. This paper describes the details of our system for the task. For feature extraction, we propose an EMD-based bag-of-feature method to exploit visual/spatial information, and utilize WordNet to expand semantic meanings of text to boost up the generalization of detectors. We also explore audio features and extract the motion cues in compressed domain for detecting concepts highly associated with audio/motion. We use Ordered Weighted Average (OWA) fusion method to combine the SVM-based multi-modal concept detection results. Experiment results show that our methods are effective.

/pdf/trecvid-2007-high-level-feature-extraction-by-mcg-ict-cas-3zcq8rdlq4.pdf

TRECVID 2007 High-Level Feature Extraction By MCG-ICT-CAS *

Unpaired image-to-image translation problem aims to model the mapping from one domain to another with unpaired training data. Current works like the well-acknowledged Cycle GAN provide a general solution for any two domains through modeling injective mappings with a symmetric structure. While in situations where two domains are asymmetric in complexity, i.e., the amount of information between two domains is different, these approaches pose problems of poor generation quality, mapping ambiguity, and model sensitivity. To address these issues, we propose Asymmetric GAN (AsymGAN) to adapt the asymmetric domains by introducing an auxiliary variable (aux) to learn the extra information for transferring from the information-poor domain to the information-rich domain, which improves the performance of state-of-the-art approaches in the following ways. First, aux better balances the information between two domains which benefits the quality of generation. Second, the imbalance of information commonly leads to mapping ambiguity, where we are able to model one-to-many mappings by tuning aux, and furthermore, our aux is controllable. Third, the training of Cycle GAN can easily make the generator pair sensitive to small disturbances and variations while our model decouples the ill-conditioned relevance of generators by injecting aux during training. We verify the effectiveness of our proposed method both qualitatively and quantitatively on asymmetric situation, label-photo task, on Cityscapes and Helen datasets, and show many applications of asymmetric image translations. In conclusion, our AsymGAN provides a better solution for unpaired image-to-image translation in asymmetric domains.

Asymmetric GAN for Unpaired Image-to-image Translation

Most of the existing scene parsing methods suffer from the serious problems of both inconsistent parsing results and object boundary shift. To tackle these issues, we first propose a Global-residual Refinement Network (GRN) through exploiting global contextual information to predict the parsing residuals and iteratively smoothen the inconsistent parsing labels. Furthermore, we propose a Localboundary Refinement Network (LRN) to learn the position-adaptive propagation coefficients so that local contextual information from neighbors can be optimally captured for refining object boundaries. Finally, we cascade the proposed two refinement networks after a fully residual convolutional neural network within a uniform framework. Extensive experiments on ADE20K and Cityscapes datasets well demonstrate the effectiveness of the two refinement methods for refining scene parsing predictions.

/pdf/global-residual-and-local-boundary-refinement-networks-for-4io78tls92.pdf

Global-residual and local-boundary refinement networks for rectifying scene parsing predictions

Realistic human action recognition in videos has been a useful yet challenging task. Video shots of same actions may present huge intra-class variations in terms of visual appearance, kinetic patterns, video shooting, and editing styles. Heterogeneous feature representations of videos pose another challenge on how to effectively handle the redundancy, complementariness and disagreement in these features. This paper proposes a localized multiple kernel learning (L-MKL) algorithm to tackle the issues above. L-MKL integrates the localized classifier ensemble learning and multiple kernel learning in a unified framework to leverage the strengths of both. The basis of L-MKL is to build multiple kernel classifiers on diverse features at subspace localities of heterogeneous representations. L-MKL integrates the discriminability of complementary features locally and enables localized MKL classifiers to deliver better performance in its own region of expertise. Specifically, L-MKL develops a locality gating model to partition the input space of heterogeneous representations to a set of localities of simpler data structure. Each locality then learns its localized optimal combination of Mercer kernels of heterogeneous features. Finally, the gating model coordinates the localized multiple kernel classifiers globally to perform action recognition. Experiments on two datasets show that the proposed approach delivers promising performance.

Localized Multiple Kernel Learning for Realistic Human Action Recognition in Videos

Effective feature extraction is a fundamental component of content-based image retrieval. Scale Invariant Feature Transform (SIFT) has been proven to be the most robust local invariant feature descriptor. However, SIFT algorithm generates hundreds of thousands of keypoints per image, and most of them comes from background. This has seriously affected the application of SIFT in real-time image retrieval. This paper addresses this problem and proposes a novel method to filter the SIFT keypoints using attention model. Based on visual attention analysis, all of the keypoints in an image are ranked with their attention saliency, and only the most distinctive keypoints will be reserved. Then we use Bag of words to efficiently index these features. Experiments demonstrate that the attention model based SIFT keypoints filtration algorithm provides significant benefits both in retrieval accuracy and matching speed.

Sheng Tang

Papers

TRECVID 2007 High-Level Feature Extraction By MCG-ICT-CAS *

Asymmetric GAN for Unpaired Image-to-image Translation

Global-residual and local-boundary refinement networks for rectifying scene parsing predictions

Localized Multiple Kernel Learning for Realistic Human Action Recognition in Videos

Attention Model Based SIFT Keypoints Filtration for Image Retrieval