计算机视觉论文-2021-07-14

本专栏是计算机视觉方向论文收集积累，时间：2021年7月14日，来源：paper digest

欢迎关注原创公众号 【计算机视觉联盟】，回复 【西瓜书手推笔记】 可获取我的机器学习纯手推笔记！

直达笔记地址：机器学习手推笔记（GitHub地址）

1, TITLE: Lifting The Convex Conjugate in Lagrangian Relaxations: A Tractable Approach for Continuous Markov Random Fields
AUTHORS: Hartmut Bauermeister ; Emanuel Laude ; Thomas M�llenhoff ; Michael Moeller ; Daniel Cremers
CATEGORY: math.OC [math.OC, cs.CV]
HIGHLIGHT: To showcase the scalability of our approach, we apply it to the stereo matching problem between two images.

2, TITLE: ST-DETR: Spatio-Temporal Object Traces Attention Detection Transformer
AUTHORS: Eslam Mohamed ; Ahmad El-Sallab
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: We propose ST-DETR, a Spatio-Temporal Transformer-based architecture for object detection from a sequence of temporal frames.

3, TITLE: PU-Flow: A Point Cloud Upsampling Networkwith Normalizing Flows
AUTHORS: AIHUA MAO et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: To address this issue, we present a novel deep learning-based model, called PU-Flow,which incorporates normalizing flows and feature interpolation techniques to produce dense points uniformly distributed on the underlying surface.

4, TITLE: Automatic Seizure Detection Using The Pulse Transit Time
AUTHORS: Eric Fiege ; Salima Houta ; Pinar Bisgin ; Rainer Surges ; Falk Howar
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this work, we present an algorithm which responds to alterations in the PTT, considering the clock drift and enabling the noninvasive monitoring of blood pressure alterations using separated sensors.

5, TITLE: Multitask Identity-Aware Image Steganography Via Minimax Optimization
AUTHORS: JIABAO CUI et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: To address this issue, we propose a framework, called Multitask Identity-Aware Image Steganography (MIAIS), to achieve direct recognition on container images without restoring secret images.

6, TITLE: Teaching Agents How to Map: Spatial Reasoning for Multi-Object Navigation
AUTHORS: Pierre Marza ; Laetitia Matignon ; Olivier Simonin ; Christian Wolf
CATEGORY: cs.CV [cs.CV, cs.LG, cs.RO]
HIGHLIGHT: We introduce supplementary supervision in the form of auxiliary tasks designed to favor the emergence of spatial perception capabilities in agents trained for a goal-reaching downstream objective.

7, TITLE: Object Tracking and Geo-localization from Street Images
AUTHORS: DANIEL WILSON et. al.
CATEGORY: cs.CV [cs.CV, cs.LG, cs.RO]
HIGHLIGHT: In this paper we present a two-stage framework that detects and geolocalizes traffic signs from low frame rate street videos.

8, TITLE: ReLLIE: Deep Reinforcement Learning for Customized Low-Light Image Enhancement
AUTHORS: Rongkai Zhang ; Lanqing Guo ; Siyu Huang ; Bihan Wen
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: To tackle these two challenges, this paper presents a novel deep reinforcement learning based method, dubbed ReLLIE, for customized low-light enhancement.

9, TITLE: Retrieve in Style: Unsupervised Facial Feature Transfer and Retrieval
AUTHORS: Min Jin Chong ; Wen-Sheng Chu ; Abhishek Kumar
CATEGORY: cs.CV [cs.CV, cs.LG]
HIGHLIGHT: We present Retrieve in Style (RIS), an unsupervised framework for fine-grained facial feature transfer and retrieval on real images.

10, TITLE: Everybody Is Unique: Towards Unbiased Human Mesh Recovery
AUTHORS: Ren Li ; Meng Zheng ; Srikrishna Karanam ; Terrence Chen ; Ziyan Wu
CATEGORY: cs.CV [cs.CV, cs.GR, cs.LG, cs.RO, stat.ML]
HIGHLIGHT: In this work, we identify this crucial gap in the current literature by presenting and discussing limitations of existing algorithms.

11, TITLE: Bayesian Atlas Building with Hierarchical Priors for Subject-specific Regularization
AUTHORS: Jian Wang ; Miaomiao Zhang
CATEGORY: cs.CV [cs.CV, eess.IV, I.2.10; I.4; I.5]
HIGHLIGHT: This paper presents a novel hierarchical Bayesian model for unbiased atlas building with subject-specific regularizations of image registration.

12, TITLE: Detect and Locate: A Face Anti-Manipulation Approach with Semantic and Noise-level Supervision
AUTHORS: CHENQI KONG et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: Herein, we propose a conceptually simple but effective method to efficiently detect forged faces in an image while simultaneously locating the manipulated regions.

13, TITLE: Dynamic Distribution of Edge Intelligence at The Node Level for Internet of Things
AUTHORS: Hawzhin Mohammed ; Tolulope A. Odetola ; Nan Guo ; Syed Rafay Hasan
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, dynamic deployment of Convolutional Neural Network (CNN) architecture is proposed utilizing only IoT-level devices.

14, TITLE: EProduct: A Million-Scale Visual Search Benchmark to Address Product Recognition Challenges
AUTHORS: Jiangbo Yuan ; An-Ti Chiang ; Wen Tang ; Antonio Haro
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: We present eProduct as a training set and an evaluation set, where the training set contains 1.3M+ listing images with titles and hierarchical category labels, for model development, and the evaluation set includes 10,000 query and 1.1 million index images for visual search evaluation.

15, TITLE: NucMM Dataset: 3D Neuronal Nuclei Instance Segmentation at Sub-Cubic Millimeter Scale
AUTHORS: ZUDI LIN et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: To tackle the challenges, we propose a novel hybrid-representation learning model that combines the merits of foreground mask, contour map, and signed distance transform to produce high-quality 3D masks.

16, TITLE: Fast and Explicit Neural View Synthesis
AUTHORS: PENGSHENG GUO et. al.
CATEGORY: cs.CV [cs.CV, cs.GR, cs.LG]
HIGHLIGHT: We propose a simple yet effective approach that is neither continuous nor implicit, challenging recent trends on view synthesis.

17, TITLE: Detect and Defense Against Adversarial Examples in Deep Learning Using Natural Scene Statistics and Adaptive Denoising
AUTHORS: Anouar Kherchouche ; Sid Ahmed Fezza ; Wassim Hamidouche
CATEGORY: cs.CV [cs.CV, cs.CR, eess.IV]
HIGHLIGHT: In this paper, we proposea framework for defending DNN classifier against ad-versarial samples.

18, TITLE: Per-Pixel Classification Is Not All You Need for Semantic Segmentation
AUTHORS: Bowen Cheng ; Alexander G. Schwing ; Alexander Kirillov
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: Following this observation, we propose MaskFormer, a simple mask classification model which predicts a set of binary masks, each associated with a single global class label prediction.

19, TITLE: This Person (Probably) Exists. Identity Membership Attacks Against GAN Generated Faces
AUTHORS: Ryan Webster ; Julien Rabin ; Loic Simon ; Frederic Jurie
CATEGORY: cs.CV [cs.CV, cs.AI]
HIGHLIGHT: In this work, we challenge the assumption that GAN faces really are novel creations, by constructing a successful membership attack of a new kind.

20, TITLE: Bidirectional Regression for Arbitrary-Shaped Text Detection
AUTHORS: Tao Sheng ; Zhouhui Lian
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: This paper presents a novel text instance expression which integrates both foreground and background information into the pipeline, and naturally uses the pixels near text boundaries as the offset starts.

21, TITLE: Scalable Surface Reconstruction with Delaunay-Graph Neural Networks
AUTHORS: Raphael Sulzer ; Loic Landrieu ; Renaud Marlet ; Bruno Vallet
CATEGORY: cs.CV [cs.CV, cs.CG]
HIGHLIGHT: We introduce a novel learning-based, visibility-aware, surface reconstruction method for large-scale, defect-laden point clouds.

22, TITLE: Deep Learning Approaches to Earth Observation Change Detection
AUTHORS: ANTONIO DI PILATO et. al.
CATEGORY: cs.CV [cs.CV, cs.AI]
HIGHLIGHT: In this paper we present two different approaches to change detection (semantic segmentation and classification) that both exploit convolutional neural networks to achieve good results, which can be further refined and used in a post-processing workflow for a large variety of applications.

23, TITLE: Learning Aesthetic Layouts Via Visual Guidance
AUTHORS: Qingyuan Zheng ; Zhuoru Li ; Adam Bargteil
CATEGORY: cs.CV [cs.CV, cs.GR, cs.MM]
HIGHLIGHT: We explore computational approaches for visual guidance to aid in creating aesthetically pleasing art and graphic design. First, we collected a dataset of art masterpieces and labeled the visual fixations with state-of-art vision models.

24, TITLE: Locally Enhanced Self-Attention: Rethinking Self-Attention As Local and Context Terms
AUTHORS: Chenglin Yang ; Siyuan Qiao ; Adam Kortylewski ; Alan Yuille
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: Therefore, we propose Locally Enhanced Self-Attention (LESA), which enhances the unary term by incorporating it with convolutions, and utilizes a fusion module to dynamically couple the unary and binary operations.

25, TITLE: CMT: Convolutional Neural Networks Meet Vision Transformers
AUTHORS: JIANYUAN GUO et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, we aim to address this issue and develop a network that can outperform not only the canonical transformers, but also the high-performance convolutional models.

26, TITLE: Cats, Not CAT Scans: A Study of Dataset Similarity in Transfer Learning for 2D Medical Image Classification
AUTHORS: Irma van den Brandt ; Floris Fok ; Bas Mulders ; Joaquin Vanschoren ; Veronika Cheplygina
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper we perform a systematic study with nine source datasets with natural or medical images, and three target medical datasets, all with 2D images.

27, TITLE: MSR-Net: Multi-Scale Relighting Network for One-to-One Relighting
AUTHORS: Sourya Dipta Das ; Nisarg A. Shah ; Saikat Dutta
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: Keeping these issues in mind, we propose the use of Stacked Deep Multi-Scale Hierarchical Network, which aggregates features from each image at different scales.

28, TITLE: MINERVAS: Massive INterior EnviRonments VirtuAl Synthesis
AUTHORS: HAOCHENG REN et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: This paper presents MINERVAS, a Massive INterior EnviRonments VirtuAl Synthesis system, to facilitate the 3D scene modification and the 2D image synthesis for various vision tasks.

29, TITLE: A Novel Deep Learning Method for Thermal to Annotated Thermal-Optical Fused Images
AUTHORS: SURANJAN GOSWAMI et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: Our method is also unique in the sense that the deep learning method we are proposing here works on the Discrete Wavelet Transform (DWT) domain instead of the gray level domain. As a part of this work, we also present a new and unique database for obtaining the region of interest in thermal images based on an existing thermal visual paired database, containing the Region of Interest on 5 different classes of data.

30, TITLE: Visual Parser: Representing Part-whole Hierarchies with Transformers
AUTHORS: Shuyang Sun* ; Xiaoyu Yue* ; Song Bai ; Philip Torr
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: This paper presents the Visual Parser (ViP) that explicitly constructs such a hierarchy with transformers.

31, TITLE: Force-in-domain GAN Inversion
AUTHORS: Guangjie Leng ; Yeku Zhu ; Zhi-Qin John Xu
CATEGORY: cs.CV [cs.CV, cs.LG, eess.IV]
HIGHLIGHT: To solve this problem, we propose a force-in-domain GAN based on the in-domain GAN, which utilizes a discriminator to force the inverted code within the latent space.

32, TITLE: Towards Building A Food Knowledge Graph for Internet of Food
AUTHORS: Weiqing Min ; Chunlin Liu ; Shuqiang Jiang
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: Scope and approach: We review the evolution of food knowledge organization, from food classification, food ontology to food knowledge graphs.

33, TITLE: CentripetalText: An Efficient Text Instance Representation for Scene Text Detection
AUTHORS: Tao Sheng ; Jie Chen ; Zhouhui Lian
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, we propose an efficient text instance representation named CentripetalText (CT), which decomposes text instances into the combination of text kernels and centripetal shifts.

34, TITLE: HAT: Hierarchical Aggregation Transformers for Person Re-identification
AUTHORS: Guowen Zhang ; Pingping Zhang ; Jinqing Qi ; Huchuan Lu
CATEGORY: cs.CV [cs.CV, cs.AI]
HIGHLIGHT: In this work, we take advantages of both CNNs and Transformers, and propose a novel learning framework named Hierarchical Aggregation Transformer (HAT) for image-based person Re-ID with high performance.

35, TITLE: Fast Batch Nuclear-norm Maximization and Minimization for Robust Domain Adaptation
AUTHORS: SHUHAO CUI et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, we investigate the prediction discriminability and diversity by studying the structure of the classification output matrix of a randomly selected data batch.

36, TITLE: Learning A Discriminant Latent Space with Neural Discriminant Analysis
AUTHORS: Mai Lan Ha ; Gianni Franchi ; Emanuel Aldea ; Volker Blanz
CATEGORY: cs.CV [cs.CV, cs.LG]
HIGHLIGHT: Inspired by Linear Discriminant Analysis (LDA), we propose an optimization called Neural Discriminant Analysis (NDA) for Deep Convolutional Neural Networks (DCNNs).

37, TITLE: 'CADSketchNet' -- An Annotated Sketch Dataset for 3D CAD Model Retrieval with Deep Neural Networks
AUTHORS: Bharadwaj Manda ; Shubham Dhayarkar ; Sai Mitheran ; V. K. Viekash ; Ramanathan Muthuganapathy
CATEGORY: cs.CV [cs.CV, cs.AI, cs.GR, cs.LG]
HIGHLIGHT: The research work presented in this paper aims at developing a dataset suitable for building a retrieval system for 3D CAD models based on deep learning.

38, TITLE: 3D Parametric Wireframe Extraction Based on Distance Fields
AUTHORS: Albert Matveev ; Alexey Artemov ; Denis Zorin ; Evgeny Burnaev
CATEGORY: cs.CV [cs.CV, cs.GR]
HIGHLIGHT: We present a pipeline for parametric wireframe extraction from densely sampled point clouds.

39, TITLE: Exploiting Image Translations Via Ensemble Self-Supervised Learning for Unsupervised Domain Adaptation
AUTHORS: Fabrizio J. Piva ; Gijs Dubbelman
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: To exploit the advantage of using multiple image translations, we propose an ensemble learning approach, where three classifiers calculate their prediction by taking as input features of different image translations, making each classifier learn independently, with the purpose of combining their outputs by sparse Multinomial Logistic Regression.

40, TITLE: Domain-Irrelevant Representation Learning for Unsupervised Domain Generalization
AUTHORS: XINGXUAN ZHANG et. al.
CATEGORY: cs.CV [cs.CV, cs.LG, cs.MM]
HIGHLIGHT: Specifically, we study a novel generalization problem called unsupervised domain generalization, which aims to learn generalizable models with unlabeled data.

41, TITLE: Deep Ranking with Adaptive Margin Triplet Loss
AUTHORS: Mai Lan Ha ; Volker Blanz
CATEGORY: cs.CV [cs.CV, cs.LG]
HIGHLIGHT: We propose a simple modification from a fixed margin triplet loss to an adaptive margin triplet loss.

42, TITLE: Region Attention and Graph Embedding Network for Occlusion Objective Class-based Micro-expression Recognition
AUTHORS: Qirong Mao ; Ling Zhou ; Wenming Zheng ; Xiuyan Shao ; Xiaohua Huang
CATEGORY: cs.CV [cs.CV, cs.AI]
HIGHLIGHT: This paper deeply investigates an interesting but unexplored challenging issue in MER, \ie, occlusion MER.

43, TITLE: Learning from Partially Overlapping Labels: Image Segmentation Under Annotation Shift
AUTHORS: Gregory Filbrandt ; Konstantinos Kamnitsas ; David Bernstein ; Alexandra Taylor ; Ben Glocker
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, we investigate and propose several strategies for learning from partially overlapping labels in the context of abdominal organ segmentation.

44, TITLE: EvoBA: An Evolution Strategy As A Strong Baseline ForBlack-Box Adversarial Attacks
AUTHORS: Andrei Ilie ; Marius Popescu ; Alin Stefanescu
CATEGORY: cs.CR [cs.CR, cs.CV, cs.LG]
HIGHLIGHT: We propose $\textbf{EvoBA}$, a black-box adversarial attack based on a surprisingly simple evolutionary search strategy.

45, TITLE: Hidden Convexity of Wasserstein GANs: Interpretable Generative Models with Closed-Form Solutions
AUTHORS: ARDA SAHINER et. al.
CATEGORY: cs.LG [cs.LG, cs.CV, eess.IV, math.OC, stat.ML]
HIGHLIGHT: In this work, we analyze the training of Wasserstein GANs with two-layer neural network discriminators through the lens of convex duality, and for a variety of generators expose the conditions under which Wasserstein GANs can be solved exactly with convex optimization approaches, or can be represented as convex-concave games.

46, TITLE: AlterSGD: Finding Flat Minima for Continual Learning By Alternative Training
AUTHORS: Zhongzhan Huang ; Mingfu Liang ; Senwei Liang ; Wei He
CATEGORY: cs.LG [cs.LG, cs.AI, cs.CV]
HIGHLIGHT: To alleviate these problems, in this paper, we propose a simple yet effective optimization method, called AlterSGD, to search for a flat minima in the loss landscape.

47, TITLE: SoftHebb: Bayesian Inference in Unsupervised Hebbian Soft Winner-take-all Networks
AUTHORS: Timoleon Moraitis ; Dmitry Toichkin ; Yansong Chua ; Qinghai Guo
CATEGORY: cs.LG [cs.LG, cs.AI, cs.CV, cs.NE, q-bio.NC]
HIGHLIGHT: Here we derive formally such a theory, based on biologically plausible but generic ANN elements.

48, TITLE: Combiner: Full Attention Transformer with Sparse Computation Cost
AUTHORS: HONGYU REN et. al.
CATEGORY: cs.LG [cs.LG, cs.CL, cs.CV]
HIGHLIGHT: Instead, we propose Combiner, which provides full attention capability in each attention head while maintaining low computation and memory complexity.

49, TITLE: On Designing Good Representation Learning Models
AUTHORS: Qinglin Li ; Bin Li ; Jonathan M Garibaldi ; Guoping Qiu
CATEGORY: cs.LG [cs.LG, cs.AI, cs.CV]
HIGHLIGHT: We present a conscience competitive learning algorithm which encourages the model to reach its MEXS whilst at the same time adheres to model smoothness prior.

50, TITLE: Generative Adversarial Learning Via Kernel Density Discrimination
AUTHORS: Abdelhak Lemkhenter ; Adam Bielski ; Alp Eren Sari ; Paolo Favaro
CATEGORY: cs.LG [cs.LG, cs.CV]
HIGHLIGHT: We introduce Kernel Density Discrimination GAN (KDD GAN), a novel method for generative adversarial learning.

51, TITLE: Scalable, Axiomatic Explanations of Deep Alzheimer's Diagnosis from Heterogeneous Data
AUTHORS: Sebastian P�lsterl ; Christina Aigner ; Christian Wachinger
CATEGORY: cs.LG [cs.LG, cs.CV, stat.ML]
HIGHLIGHT: We propose Shapley Value Explanation of Heterogeneous Neural Networks (SVEHNN) for explaining the Alzheimer's diagnosis made by a DNN from the 3D point cloud of the neuroanatomy and tabular biomarkers.

52, TITLE: Kit-Net: Self-Supervised Learning to Kit Novel 3D Objects Into Novel 3D Cavities
AUTHORS: SHIVIN DEVGON et. al.
CATEGORY: cs.RO [cs.RO, cs.AI, cs.CV]
HIGHLIGHT: We present Kit-Net, a framework for kitting previously unseen 3D objects into cavities given depth images of both the target cavity and an object held by a gripper in an unknown initial orientation.

53, TITLE: Combining 3D Image and Tabular Data Via The Dynamic Affine Feature Map Transform
AUTHORS: Sebastian P�lsterl ; Tom Nuno Wolf ; Christian Wachinger
CATEGORY: eess.IV [eess.IV, cs.CV, cs.LG]
HIGHLIGHT: We introduce the Dynamic Affine Feature Map Transform (DAFT), a general-purpose module for CNNs that dynamically rescales and shifts the feature maps of a convolutional layer, conditional on a patient's tabular clinical information.

54, TITLE: Attention Based CNN-LSTM Network for Pulmonary Embolism Prediction on Chest Computed Tomography Pulmonary Angiograms
AUTHORS: SUDHIR SUMAN et. al.
CATEGORY: eess.IV [eess.IV, cs.CV]
HIGHLIGHT: In this study we propose a two-stage attention-based CNN-LSTM network for predicting PE, its associated type (chronic, acute) and corresponding location (leftsided, rightsided or central) on computed tomography (CT) examinations.

55, TITLE: Attention-Guided Progressive Neural Texture Fusion for High Dynamic Range Image Restoration
AUTHORS: JIE CHEN et. al.
CATEGORY: eess.IV [eess.IV, cs.CV]
HIGHLIGHT: In this work, we propose an Attention-guided Progressive Neural Texture Fusion (APNT-Fusion) HDR restoration model which aims to address these issues within one framework.

56, TITLE: Detecting When Pre-trained NnU-Net Models Fail Silently for Covid-19
AUTHORS: CAMILA GONZALEZ et. al.
CATEGORY: eess.IV [eess.IV, cs.CV, cs.LG]
HIGHLIGHT: We propose a lightweight OOD detection method that exploits the Mahalanobis distance in the feature space.