The central building block of convolutional neural networks (CNNs) is the convolution operator, which enables networks to construct informative features by fusing both spatial and channel-wise information within local receptive fields at each layer. A broad range of prior research has investigated the spatial component of this relationship, seeking to strengthen the representational power of a CNN by enhancing the quality of spatial encodings throughout its feature hierarchy. In this work, we focus instead on the channel relationship and propose a novel architectural unit, which we term the “Squeeze-and-Excitation” (SE) block, that adaptively recalibrates channel-wise feature responses by explicitly modelling interdependencies between channels. We show that these blocks can be stacked together to form SENet architectures that generalise extremely effectively across different datasets. We further demonstrate that SE blocks bring significant improvements in performance for existing state-of-the-art CNNs at slight additional computational cost. Squeeze-and-Excitation Networks formed the foundation of our ILSVRC 2017 classification submission which won first place and reduced the top-5 error to 2.251 percent, surpassing the winning entry of 2016 by a relative improvement of ${\sim }$ ∼ 25 percent. Models and code are available at https://github.com/hujie-frank/SENet .

/pdf/squeeze-and-excitation-networks-50icgwqgeu.pdf

Squeeze-and-Excitation Networks

Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

We propose Convolutional Block Attention Module (CBAM), a simple yet effective attention module for feed-forward convolutional neural networks. Given an intermediate feature map, our module sequentially infers attention maps along two separate dimensions, channel and spatial, then the attention maps are multiplied to the input feature map for adaptive feature refinement. Because CBAM is a lightweight and general module, it can be integrated into any CNN architectures seamlessly with negligible overheads and is end-to-end trainable along with base CNNs. We validate our CBAM through extensive experiments on ImageNet-1K, MS~COCO detection, and VOC~2007 detection datasets. Our experiments show consistent improvements in classification and detection performances with various models, demonstrating the wide applicability of CBAM. The code and models will be publicly available.

CBAM: Convolutional Block Attention Module

There are a huge number of features which are said to improve Convolutional Neural Network (CNN) accuracy. Practical testing of combinations of such features on large datasets, and theoretical justification of the result, is required. Some features operate on certain models exclusively and for certain problems exclusively, or only for small-scale datasets; while some features, such as batch-normalization and residual-connections, are applicable to the majority of models, tasks, and datasets. We assume that such universal features include Weighted-Residual-Connections (WRC), Cross-Stage-Partial-connections (CSP), Cross mini-Batch Normalization (CmBN), Self-adversarial-training (SAT) and Mish-activation. We use new features: WRC, CSP, CmBN, SAT, Mish activation, Mosaic data augmentation, CmBN, DropBlock regularization, and CIoU loss, and combine some of them to achieve state-of-the-art results: 43.5% AP (65.7% AP50) for the MS COCO dataset at a realtime speed of ~65 FPS on Tesla V100. Source code is at this https URL

YOLOv4: Optimal Speed and Accuracy of Object Detection

The central building block of convolutional neural networks (CNNs) is the convolution operator, which enables networks to construct informative features by fusing both spatial and channel-wise information within local receptive fields at each layer. A broad range of prior research has investigated the spatial component of this relationship, seeking to strengthen the representational power of a CNN by enhancing the quality of spatial encodings throughout its feature hierarchy. In this work, we focus instead on the channel relationship and propose a novel architectural unit, which we term the "Squeeze-and-Excitation" (SE) block, that adaptively recalibrates channel-wise feature responses by explicitly modelling interdependencies between channels. We show that these blocks can be stacked together to form SENet architectures that generalise extremely effectively across different datasets. We further demonstrate that SE blocks bring significant improvements in performance for existing state-of-the-art CNNs at slight additional computational cost. Squeeze-and-Excitation Networks formed the foundation of our ILSVRC 2017 classification submission which won first place and reduced the top-5 error to 2.251%, surpassing the winning entry of 2016 by a relative improvement of ~25%. Models and code are available at this https URL.

Despite the continuous advances in local stereo matching for years, most efforts are on developing robust cost computation and aggregation methods. Little attention has been seriously paid to the disparity refinement. In this work, we study weighted median filtering for disparity refinement. We discover that with this refinement, even the simple box filter aggregation achieves comparable accuracy with various sophisticated aggregation methods (with the same refinement). This is due to the nice weighted median filtering properties of removing outlier error while respecting edges/structures. This reveals that the previously overlooked refinement can be at least as crucial as aggregation. We also develop the first constant time algorithm for the previously time-consuming weighted median filter. This makes the simple combination ``box aggregation + weighted median'' an attractive solution in practice for both speed and accuracy. As a byproduct, the fast weighted median filtering unleashes its potential in other applications that were hampered by high complexities. We show its superiority in various applications such as depth up sampling, clip-art JPEG artifact removal, and image stylization.

/pdf/constant-time-weighted-median-filtering-for-stereo-matching-paso4be6cd.pdf

Constant Time Weighted Median Filtering for Stereo Matching and Beyond

We propose a method for intrinsic image decomposition based on retinex theory and texture analysis. While most previous methods approach this problem by analyzing local gradient properties, our technique additionally identifies distant pixels with the same reflectance through texture analysis, and uses these nonlocal reflectance constraints to significantly reduce ambiguity in decomposition. We formulate the decomposition problem as the minimization of a quadratic function which incorporates both the retinex constraint and our nonlocal texture constraint. This optimization can be solved in closed form with the standard conjugate gradient algorithm. Extensive experimentation with comparisons to previous techniques validate our method in terms of both decomposition accuracy and runtime efficiency.

A Closed-Form Solution to Retinex with Nonlocal Texture Constraints

In this paper, we solve the 3D fluid dynamics problem in a complex environment by taking advantage of the parallelism and programmability of GPU. In difference from other methods, innovation is made in two aspects. Firstly, more general boundary conditions could be processed on GPU in our method. By the method, we generate the boundary from a 3D scene with solid clipping, making the computation run on GPU despite of the complexity of the whole geometry scene. Then by grouping the voxels into different types according to their positions relative to the obstacles and locating the voxel that determines the value of the current voxel, we modify the values on the boundaries according to the boundary conditions. Secondly, more compact structure in data packing with flat 3D textures is designed at the fragment processing level to enhance parallelism and reduce execution passes. The scalar variables including density and temperature are packed into four channels of texels to accelerate the computation of 3D Navier-Stokes equations (NSEs). The test results prove the efficiency of our method, and as a result, it is feasible to run middle-scale problems of 3D fluid dynamics in an interactive speed for more general environment with complex geometry on PC platform.

/pdf/real-time-3d-fluid-simulation-on-gpu-with-complex-obstacles-2mbbvjh6ah.pdf

Enhua Wu

Papers

Squeeze-and-Excitation Networks

Squeeze-and-Excitation Networks

Constant Time Weighted Median Filtering for Stereo Matching and Beyond

A Closed-Form Solution to Retinex with Nonlocal Texture Constraints

Real-time 3D fluid simulation on GPU with complex obstacles