A key challenge in video enhancement and action recognition is to fuse useful information from neighboring frames. Recent works suggest establishing accurate correspondences between neighboring frames before fusing temporal information. However, the generated results heavily depend on the quality of correspondence estimation. In this paper, we propose a more robust solution: \emph{sampling and fusing multi-level features} across neighborhood frames to generate the results. Based on this idea, we introduce a new module to improve the capability of 3D convolution, namely, learnable sampling 3D convolution (\emph{LS3D-Conv}). We add learnable 2D offsets to 3D convolution which aims to sample locations on spatial feature maps across frames. The offsets can be learned for specific tasks. The \emph{LS3D-Conv} can flexibly replace 3D convolution layers in existing 3D networks and get new architectures, which learns the sampling at multiple feature levels. The experiments on video interpolation, video super-resolution, video denoising, and action recognition demonstrate the effectiveness of our approach.
Inria（法国国家信息与自动化研究所）；WILLOW（威洛实验室）；Departement d’Informatique de l’École Normale Supérieure（巴黎高等师范学院信息学系）；PSL Research University（巴黎文理研究大学）；Czech Institute of Informatics, Robotics and Cybernetics（捷克理工大学信息学、机器人学与控制论研究所）；Czech Technical University in Prague（布拉格捷克理工大学）
Current methods for video analysis often extract frame-level features using pre-trained convolutional neural networks (CNNs). Such features are then aggregated over time e.g., by simple temporal averaging or more sophisticated recurrent neural networks such as long short-term memory (LSTM) or gated recurrent units (GRU). In this work we revise existing video representations and study alternative methods for temporal aggregation. We first explore clustering-based aggregation layers and propose a two-stream architecture aggregating audio and visual features. We then introduce a learnable non-lin…
Existing convolution techniques in artificial neural networks suffer from huge computation complexity, while the biological neural network works in a much more powerful yet efficient way. Inspired by the biological plasticity of dendritic topology and synaptic strength, our method, Learnable Heterogeneous Convolution, realizes joint learning of kernel shape and weights, which unifies existing handcrafted convolution techniques in a data-driven way. A model based on our method can converge with structural sparse weights and then be accelerated by devices of high parallelism. In the experiments,…
Brain surface analysis is essential to neuroscience, however, the complex geometry of the brain cortex hinders computational methods for this task. The difficulty arises from a discrepancy between 3D imaging data, which is represented in Euclidean space, and the non-Euclidean geometry of the highly-convoluted brain surface. Recent advances in machine learning have enabled the use of neural networks for non-Euclidean spaces. These facilitate the learning of surface data, yet pooling strategies often remain constrained to a single fixed-graph. This paper proposes a new learnable graph pooling me…
Harbin Institute of Technology（哈尔滨工业大学）；Peng Cheng Laboratory（鹏城实验室）；Baidu Inc.（百度公司）；Nankai University（南开大学）
Most convolutional network (CNN)-based inpainting methods adopt standard convolution to indistinguishably treat valid pixels and holes, making them limited in handling irregular holes and more likely to generate inpainting results with color discrepancy and blurriness. Partial convolution has been suggested to address this issue, but it adopts handcrafted feature re-normalization, and only considers forward mask-updating. In this paper, we present a learnable attention map module for learning feature renormalization and mask-updating in an end-to-end manner, which is effective in adapting to i…
Moscow Institute of Physics and Technology（莫斯科物理技术学院）
The article describes a system for image recognition using deep convolutional neural networks. Modified network architecture is proposed that focuses on improving convergence and reducing training complexity. The filters in the first layer of the network are constrained to fit the Gabor function. The parameters of Gabor functions are learnable and are updated by standard backpropagation techniques. The system was implemented on Python, tested on several datasets and outperformed the common convolutional networks.
Dilated Convolution with Learnable Spacings: beyond bilinear interpolation
Artificial and Natural Intelligence Toulouse Institute (ANITI)（图卢兹人工智能与自然智能研究所）；CerCo UMR 5549, CNRS（法国国家科学研究中心CerCo UMR 5549实验室）；Université Toulouse III（图卢兹第三大学）；IRIT, CNRS（法国国家科学研究中心图卢兹信息学研究所）；Toulouse INP（图卢兹国立综合理工学院）
Harbin Institute of Technology（哈尔滨工业大学）；Peng Cheng Lab（鹏城实验室）；DAMO Academy at Alibaba Group（阿里巴巴达摩院）
For image inpainting, the convolutional neural networks (CNN) in previous methods often adopt standard convolutional operator, which treats valid pixels and holes indistinguishably. As a result, they are limited in handling irregular holes and tend to produce color-discrepant and blurry inpainting result. Partial convolution (PConv) copes with this issue by conducting masked convolution and feature re-normalization conditioned only on valid pixels, but the mask-updating is handcrafted and independent with image structural information. In this paper, we present an edge-guided learnable bidirect…
University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）；Purdue University（普渡大学）
We propose learnable polyphase sampling (LPS), a pair of learnable down/upsampling layers that enable truly shift-invariant and equivariant convolutional networks. LPS can be trained end-to-end from data and generalizes existing handcrafted downsampling layers. It is widely applicable as it can be integrated into any convolutional network by replacing down/upsampling layers. We evaluate LPS on image classification and semantic segmentation. Experiments show that LPS is on-par with or outperforms existing methods in both performance and shift consistency. For the first time, we achieve true shi…
Fuzhou University（福州大学）；University of Vienna（维也纳大学）
In practical applications, multi-view data depicting objectives from assorted perspectives can facilitate the accuracy increase of learning algorithms. However, given multi-view data, there is limited work for learning discriminative node relationships and graph information simultaneously via graph convolutional network that has drawn the attention from considerable researchers in recent years. Most of existing methods only consider the weighted sum of adjacency matrices, yet a joint neural network of both feature and graph fusion is still under-explored. To cope with these issues, this paper …
Institute of Computing Technology, CAS（中国科学院计算技术研究所）；University of Chinese Academy of Sciences（中国科学院大学）；CAS Center for Excellence in Brain Science and Intelligence Technology（中国科学院脑科学与智能技术卓越创新中心）
Benefitted from its great success on many tasks, deep learning is increasingly used on low-computational-cost devices, e.g. smartphone, embedded devices, etc. To reduce the high computational and memory cost, in this work, we propose a fully learnable group convolution module (FLGC for short) which is quite efficient and can be embedded into any deep neural networks for acceleration. Specifically, our proposed method automatically learns the group structure in the training stage in a fully end-to-end manner, leading to a better structure than the existing pre-defined, two-steps, or iterative s…
University of Toronto（多伦多大学）；Uber Advanced Technologies Group（优步先进技术集团）
Point clouds are the native output of many real-world 3D sensors. To borrow the success of 2D convolutional network architectures, a majority of popular 3D perception models voxelize the points, which can result in a loss of local geometric details that cannot be recovered. In this paper, we propose a novel learnable convolution layer for processing 3D point cloud data directly. Instead of discretizing points into fixed voxels, we deform our learnable 3D filters to match with the point cloud shape. We propose to combine voxelized backbone networks with our deformable filter layer at 1) the net…
Department of Computer Science Norwegian University of Science and Technology（挪威科技大学计算机科学系）；Norwegian University of Science and Technology（挪威科技大学）
A regular convolution layer applying a filter in the same way over known and unknown areas causes visual artifacts in the inpainted image. Several studies address this issue with feature re-normalization on the output of the convolution. However, these models use a significant amount of learnable parameters for feature re-normalization, or assume a binary representation of the certainty of an output. We propose (layer-wise) feature imputation of the missing input values to a convolution. In contrast to learned feature re-normalization, our method is efficient and introduces a minimal number of…
Geometric variations like rotation, scaling, and viewpoint changes pose a significant challenge to visual understanding. One common solution is to directly model certain intrinsic structures, e.g., using landmarks. However, it then becomes non-trivial to build effective deep models, especially when the underlying non-Euclidean grid is irregular and coarse. Recent deep models using graph convolutions provide an appropriate framework to handle such non-Euclidean data, but many of them, particularly those based on global graph Laplacians, lack expressiveness to capture local features required for…
Microsoft Research Asia（微软亚洲研究院）；Xi’an Jiaotong University（西安交通大学）；Microsoft Research（微软研究院）
Image enhancement aims at improving the aesthetic visual quality of photos by retouching the color and tone, and is an essential technology for professional digital photography. Recent years deep learning-based image enhancement algorithms have achieved promising performance and attracted increasing popularity. However, typical efforts attempt to construct a uniform enhancer for all pixels' color transformation. It ignores the pixel differences between different content (e.g., sky, ocean, etc.) that are significant for photographs, causing unsatisfactory results. In this paper, we propose a no…
Structural Kolmogorov-Arnold Convolutions: Learnable Function on the Values or the Filter Shape as Parameter-Efficient Alternative to Per-Edge Convolutional KANs
Istituto Italiano di Tecnologia（意大利技术研究院）；Department of Informatics, Bioengineering, Robotics and Systems Engineering (DIBRIS), University of Genoa（热那亚大学信息学、生物工程、机器人与系统工程系）；Department of Theoretical and Applied Sciences (DISTA), eCampus University（eCampus大学理论与应用科学系）；Department of Intelligent Software Systems and Technologies, School of Computer Science and Artificial Intelligence, V. N. Karazin Kharkiv National University（哈尔科夫国立大学计算机科学与人工智能学院智能软件系统与技术系）；Department of Information Engineering, Marche Polytechnic University（马尔凯理工大学信息工程系）
The Hong Kong Polytechnic University（香港理工大学）；Xidian University（西安电子科技大学）
2D convolution (Conv2d), which is responsible for extracting features from the input image, is one of the key modules of a convolutional neural network (CNN). However, Conv2d is vulnerable to image corruptions and adversarial samples. It is an important yet rarely investigated problem that whether we can design a more robust alternative of Conv2d for more reliable feature extraction. In this paper, inspired by the recently developed learnable sparse transform that learns to convert the CNN features into a compact and sparse latent space, we design a novel building block, denoted by RConv-MK, t…
Artificial and Natural Intelligence Toulouse Institute (ANITI)（图卢兹人工智能与自然智能研究所（ANITI））；Université de Toulouse（图卢兹大学）；IRIT（图卢兹计算机研究所）；CNRS（法国国家科学研究中心）；Toulouse INP（图卢兹国立理工学院）；UT3（图卢兹第三大学）；CerCo UMR 5549（认知与脑科学研究中心（UMR 5549））
Convolutional neural networks are able to perform a hierarchical learning process starting with local features. However, a limited attention is paid to enhancing such elementary level features like edges. We propose and evaluate two wavelet-based edge feature enhancement methods to preprocess the input images to convolutional neural networks. The first method develops feature enhanced representations by decomposing the input images using wavelet transform and limited reconstructing subsequently. The second method develops such feature enhanced inputs to the network using local modulus maxima o…
We propose a computationally efficient architecture that learns to segment lesions from CT images of the liver. The proposed architecture uses bilinear interpolation with sub-pixel convolution at the last layer to upscale the course feature in bottle neck architecture. Since bilinear interpolation and sub-pixel convolution do not have any learnable parameter, our overall model is faster and occupies less memory footprint than the traditional U-net. We evaluate our proposed architecture on the highly competitive dataset of 2017 Liver Tumor Segmentation (LiTS) Challenge. Our method achieves comp…
The Pennsylvania State University（宾夕法尼亚州立大学）；Carnegie Mellon University（卡内基梅隆大学）；Facebook（脸书）
Cervical cancer is the second most prevalent cancer affecting women today. As the early detection of cervical carcinoma relies heavily upon screening and pre-clinical testing, digital cervicography has great potential as a primary or auxiliary screening tool, especially in low-resource regions due to its low cost and easy access. Although an automated cervical dysplasia detection system has been desirable, traditional fully-supervised training of such systems requires large amounts of annotated data which are often labor-intensive to collect. To alleviate the need for much manual annotation, w…
FreeTransformSR: Efficient Lightweight Image Super-Resolution via Free Low-Rank Learnable Transform
Chinese Academy of Sciences Changchun Institute of Optics Fine Mechanics and Physics（中国科学院长春光学精密机械与物理研究所）；University of the Chinese Academy of Sciences（中国科学院大学）
State Key Laboratory of Automotive Safety and Energy（汽车安全与节能国家重点实验室）；School of Vehicle and Mobility（车辆与动力学院）；Tsinghua University（清华大学）
3D object detection plays a crucial role in environmental perception for autonomous vehicles, which is the prerequisite of decision and control. This paper analyses partition-based methods' inherent drawbacks. In the partition operation, a single instance such as a pedestrian is sliced into several pieces, which we call it the partition effect. We propose the Spatial-Attention Graph Convolution (S-AT GCN), forming the Feature Enhancement (FE) layers to overcome this drawback. The S-AT GCN utilizes the graph convolution and the spatial attention mechanism to extract local geometrical structure …
Recently, deep learning-based image enhancement algorithms achieved state-of-the-art (SOTA) performance on several publicly available datasets. However, most existing methods fail to meet practical requirements either for visual perception or for computation efficiency, especially for high-resolution images. In this paper, we propose a novel real-time image enhancer via learnable spatial-aware 3-dimentional lookup tables(3D LUTs), which well considers global scenario and local spatial information. Specifically, we introduce a light weight two-head weight predictor that has two outputs. One is …
Skolkovo Institute of Science and Technology（斯科尔科沃科学技术学院）；FSBI National Medical Research Radiological Center of the Ministry of Health of the Russian Federation（俄罗斯联邦卫生部国家医学研究放射中心）
Segmentation of certain hollow organs, such as the bladder, is especially hard to automate due to their complex geometry, vague intensity gradients in the soft tissues, and a tedious manual process of the data annotation routine. Yet, accurate localization of the walls and the cancer regions in the radiologic images of such organs is an essential step in oncology. To address this issue, we propose a new class of hollow kernels that learn to 'mimic' the contours of the segmented organ, effectively replicating its shape and structural complexity. We train a series of the U-Net-like neural networ…
Huazhong University of Science and Technology（华中科技大学）；Tencent AI Lab（腾讯AI实验室）
In this paper, we propose to estimate 3D hand pose by recovering the 3D coordinates of joints in a group-wise manner, where less-related joints are automatically categorized into different groups and exhibit different features. This is different from the previous methods where all the joints are considered holistically and share the same feature. The benefits of our method are illustrated by the principle of multi-task learning (MTL), i.e., by separating less-related joints into different groups (as different tasks), our method learns different features for each of them, therefore efficiently …
Monocular Depth Estimation via Neural Network with Learnable Algebraic Group and Ring Structures
Chengdu Institute of Computer Applications, Chinese Academy of Sciences（中国科学院成都计算机应用研究所）；School of Computer Science and Technology, University of Chinese Academy of Sciences（中国科学院大学计算机科学与技术学院）；School of Computing and Artificial Intelligence, Southwest Jiaotong University（西南交通大学计算机科学与人工智能学院）；School of Mechanical Engineering, Southwest Jiaotong University（西南交通大学机械工程学院）；School of Computer Science and Engineering, University of Electronic Science and Technology of China（电子科技大学计算机科学与工程学院）
National Yang Ming Chiao Tung University（阳明交通大学）；University at Albany - SUNY（奥尔巴尼大学 - 纽约州立大学）；University at Buffalo（布法罗大学）
Pooling is a simple but essential layer in modern deep CNN architectures for feature aggregation and extraction. Typical CNN design focuses on the conv layers and activation functions, while leaving the pooling layers with fewer options. We introduce the Learning Discrete Wavelet Pooling (LDW-Pooling) that can be applied universally to replace standard pooling operations to better extract features with improved accuracy and efficiency. Motivated from the wavelet theory, we adopt the low-pass (L) and high-pass (H) filters horizontally and vertically for pooling on a 2D feature map. Feature sign…
Xidian University（西安电子科技大学）；Universitat Autònoma de Barcelona（巴塞罗那自治大学）；Northwestern Polytechnical University（西北工业大学）
In this paper, we propose a deformable convolution-based generative adversarial network (DCNGAN) for perceptual quality enhancement of compressed videos. DCNGAN is also adaptive to the quantization parameters (QPs). Compared with optical flows, deformable convolutions are more effective and efficient to align frames. Deformable convolutions can operate on multiple frames, thus leveraging more temporal information, which is beneficial for enhancing the perceptual quality of compressed videos. Instead of aligning frames in a pairwise manner, the deformable convolution can process multiple frames…
This paper presents a new approach for the recognition of elements in floor plan layouts. Besides of elements with common shapes, we aim to recognize elements with irregular shapes such as circular rooms and inclined walls. Furthermore, the reduction of noise in the semantic segmentation of the floor plan is on demand. To this end, we propose direction-aware, learnable, additive kernels in the application of both the context module and common convolutional blocks. We apply them for high performance of elements with both common and irregular shapes. Besides, an adversarial network with two disc…
In this paper, we propose a novel convolutional neural network (CNN) architecture considering both local and global features for image enhancement. Most conventional image enhancement methods, including Retinex-based methods, cannot restore lost pixel values caused by clipping and quantizing. CNN-based methods have recently been proposed to solve the problem, but they still have a limited performance due to network architectures not handling global features. To handle both local and global features, the proposed architecture consists of three networks: a local encoder, a global encoder, and a …
Shandong Normal University（山东师范大学）；Nanjing University of Science and Technology（南京理工大学）
Low-light image enhancement is a classical computer vision problem aiming to recover normal-exposure images from low-light images. However, convolutional neural networks commonly used in this field are good at sampling low-frequency local structural features in the spatial domain, which leads to unclear texture details of the reconstructed images. To alleviate this problem, we propose a novel module using the Fourier coefficients, which can recover high-quality texture details under the constraint of semantics in the frequency phase and supplement the spatial domain. In addition, we design a s…
In this paper, we propose a novel convolutional neural network (CNN) that never causes checkerboard artifacts, for image enhancement. In research fields of image-to-image translation problems, it is well-known that images generated by usual CNNs are distorted by checkerboard artifacts which mainly caused in forward-propagation of upsampling layers. However, checkerboard artifacts in image enhancement have never been discussed. In this paper, we point out that applying U-Net based CNNs to image enhancement causes checkerboard artifacts. In contrast, the proposed network that contains fixed conv…
