计算机视觉论文-20210223_learning neural network subspaces

本专栏是计算机视觉方向论文收集积累，时间：2021年2月23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: Explainers in The Wild: Making Surrogate Explainers Robust to Distortions Through Perception
AUTHORS: Alexander Hepburn ; Raul Santos-Rodriguez
CATEGORY: cs.CV [cs.CV, stat.ML]
HIGHLIGHT: In this paper we propose a methodology to evaluate the effect of distortions in explanations by embedding perceptual distances that tailor the neighbourhoods used to training surrogate explainers.

2, TITLE: Deep Learning for Robust Motion Segmentation with Non-Static Cameras
AUTHORS: Markus Bosch
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: This work proposes a new end-to-end DCNN based approach for motion segmentation, especially for video sequences captured with such non-static cameras, called MOSNET.

3, TITLE: Escaping Poor Local Minima in Large Scale Robust Estimation
AUTHORS: Huu Le ; Christopher Zach
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, we introduce two novel approaches for robust parameter estimation.

4, TITLE: Approximation of Dilation-based Spatial Relations to Add Structural Constraints in Neural Networks
AUTHORS: Mateus Riva ; Pietro Gori ; Florian Yger ; Roberto Cesar ; Isabelle Bloch
CATEGORY: cs.CV [cs.CV, cs.AI]
HIGHLIGHT: We propose to approximate dilations using convolutions based on a kernel equal to the structuring element.

5, TITLE: Contour Loss for Instance Segmentation Via K-step Distance Transformation Image
AUTHORS: Xiaolong Guo ; Xiaosong Lan ; Kunfeng Wang ; Shuxiao Li
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: To cope with this problem, we draw on the idea of contour matching based on distance transformation image and propose a novel loss function, called contour loss.

6, TITLE: GMLight: Lighting Estimation Via Geometric Distribution Approximation
AUTHORS: FANGNENG ZHAN et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: This paper presents Geometric Mover's Light (GMLight), a lighting estimation framework that employs a regression network and a generative projector for effective illumination estimation.

7, TITLE: Semantically Meaningful Class Prototype Learning for One-Shot Image Semantic Segmentation
AUTHORS: TAO CHEN et. al.
CATEGORY: cs.CV [cs.CV, cs.MM]
HIGHLIGHT: In this paper, we propose to leverage the multi-class label information during the episodic training.

8, TITLE: Direct Estimation of Appearance Models for Segmentation
AUTHORS: Jeova F. S. Rocha Neto ; Pedro Felzenszwalb ; Marilyn Vazquez
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: We describe a novel approach for estimating appearance models directly from an image, without explicit consideration of the pixels that make up each region.

9, TITLE: Concealed Object Detection
AUTHORS: Deng-Ping Fan ; Ge-Peng Ji ; Ming-Ming Cheng ; Ling Shao
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: We present the first systematic study on concealed object detection (COD), which aims to identify objects that are "perfectly" embedded in their background. To better understand this task, we collect a large-scale dataset, called COD10K, which consists of 10,000 images covering concealed objects in diverse real-world scenarios from 78 object categories.

10, TITLE: Adaptable Deformable Convolutions for Semantic Segmentation of Fisheye Images in Autonomous Driving Systems
AUTHORS: Cl�ment Playout ; Ola Ahmad ; Freddy Lecue ; Farida Cheriet
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: This work demonstrates that a CNN trained on standard images can be readily adapted to fisheye images, which is crucial in real-world applications where time-consuming real-time data transformation must be avoided.

11, TITLE: Image Classification Using CNN for Traffic Signs in Pakistan
AUTHORS: Abdul Azeem Sikander ; Hamza Ali
CATEGORY: cs.CV [cs.CV, cs.AI]
HIGHLIGHT: In this work to increase the accuracy, more dataset was collected to increase the size of images in every class in the data set.

12, TITLE: Transformer Is All You Need: Multimodal Multitask Learning with A Unified Transformer
AUTHORS: Ronghang Hu ; Amanpreet Singh
CATEGORY: cs.CV [cs.CV, cs.CL]
HIGHLIGHT: We propose UniT, a Unified Transformer model to simultaneously learn the most prominent tasks across different domains, ranging from object detection to language understanding and multimodal reasoning.

13, TITLE: Camera Calibration with Pose Guidance
AUTHORS: Yuzhuo Ren ; Feng Hu
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: To resolve above issues, we propose a calibration system called Calibration with Pose Guidance to improve calibration accuracy, reduce calibration variance among different users or different trials of the same person.

14, TITLE: Attention Models for Point Clouds in Deep Learning: A Survey
AUTHORS: Xu Wang ; Yi Jin ; Yigang Cen ; Tao Wang ; Yidong Li
CATEGORY: cs.CV [cs.CV, cs.AI]
HIGHLIGHT: In this paper, our ultimate goal is to provide a comprehensive overview of the point clouds feature representation which uses attention models.

15, TITLE: Towards Accurate and Compact Architectures Via Neural Architecture Transformer
AUTHORS: YONG GUO et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: To this end, we have proposed a Neural Architecture Transformer (NAT) method which casts the optimization problem into a Markov Decision Process (MDP) and seeks to replace the redundant operations with more efficient operations, such as skip or null connection.

16, TITLE: Probabilistic Vehicle Reconstruction Using A Multi-Task CNN
AUTHORS: Max Coenen ; Franz Rottensteiner
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, we present a probabilistic approach for shape-aware 3D vehicle reconstruction from stereo images that leverages the outputs of a novel multi-task CNN.

17, TITLE: PCB-Fire: Automated Classification and Fault Detection in PCB
AUTHORS: Tejas Khare ; Vaibhav Bahel ; Anuradha C. Phadke
CATEGORY: cs.CV [cs.CV, eess.IV]
HIGHLIGHT: The authors present a novel solution for detecting missing components and classifying them in a resourceful manner.

18, TITLE: Transferable Visual Words: Exploiting The Semantics of Anatomical Patterns for Self-supervised Learning
AUTHORS: Fatemeh Haghighi ; Mohammad Reza Hosseinzadeh Taher ; Zongwei Zhou ; Michael B. Gotway ; Jianming Liang
CATEGORY: cs.CV [cs.CV, eess.IV]
HIGHLIGHT: This paper introduces a new concept called "transferable visual words" (TransVW), aiming to achieve annotation efficiency for deep learning in medical image analysis.

19, TITLE: Hard-Attention for Scalable Image Classification
AUTHORS: Athanasios Papadopoulos ; Pawe? Korus ; Nasir Memon
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: We propose a novel architecture, TNet, which traverses an image pyramid in a top-down fashion, visiting only the most informative regions along the way.

20, TITLE: MetaDelta: A Meta-Learning System for Few-shot Image Classification
AUTHORS: Yudong Chen ; Chaoyu Guan ; Zhikun Wei ; Xin Wang ; Wenwu Zhu
CATEGORY: cs.CV [cs.CV, cs.AI]
HIGHLIGHT: In this paper, we propose MetaDelta, a novel practical meta-learning system for the few-shot image classification.

21, TITLE: Three Dimensional Unique Identifier Based Automated Georeferencing and Coregistration of Point Clouds in Underground Environment
AUTHORS: Sarvesh Kumar Singh ; Bikram Pratap Banerjee ; Simit Raval
CATEGORY: cs.CV [cs.CV, I.4.9]
HIGHLIGHT: This study aims at overcoming these practical challenges in underground or indoor laser scanning.

22, TITLE: Learning Deep Features for Shape Correspondence with Domain Invariance
AUTHORS: Praful Agrawal ; Ross T. Whitaker ; Shireen Y. Elhabian
CATEGORY: cs.CV [cs.CV, cs.LG]
HIGHLIGHT: This paper proposes an automated feature learning approach, using deep convolutional neural networks to extract correspondence-friendly features from shape ensembles.

23, TITLE: CheXseg: Combining Expert Annotations with DNN-generated Saliency Maps for X-ray Segmentation
AUTHORS: Soham Gadgil ; Mark Endo ; Emily Wen ; Andrew Y. Ng ; Pranav Rajpurkar
CATEGORY: cs.CV [cs.CV, cs.AI, cs.LG]
HIGHLIGHT: In this work, we propose a method that combines the high quality of pixel-level expert annotations with the scale of coarse DNN-generated saliency maps for training multi-label semantic segmentation models.

24, TITLE: Scalable Balanced Training of Conditional Generative Adversarial Neural Networks on Image Data
AUTHORS: Massimiliano Lupo Pasini ; Vittorio Gabbi ; Junqi Yin ; Simona Perotto ; Nouamane Laanait
CATEGORY: cs.CV [cs.CV, cs.AI, I.2.10; I.2.11; I.5.1; I.6.5]
HIGHLIGHT: We propose a distributed approach to train deep convolutional generative adversarial neural network (DC-CGANs) models.

25, TITLE: A Deep Learning-based Method to Extract Lumen and Media-Adventitia in Intravascular Ultrasound Images
AUTHORS: FUBAO ZHU et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, we aim to develop a deep learning-based method using an encoder-decoder deep architecture to automatically extract both lumen and MA border.

26, TITLE: A Hierarchical Conditional Random Field-based Attention Mechanism Approach for Gastric Histopathology Image Classification
AUTHORS: YIXIN LI et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, to accomplish the tasks of GHIC superiorly and to assist pathologists in clinical diagnosis, an intelligent Hierarchical Conditional Random Field based Attention Mechanism (HCRF-AM) model is proposed.

27, TITLE: Unsupervised Medical Image Alignment with Curriculum Learning
AUTHORS: Mihail Burduja ; Radu Tudor Ionescu
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: We explore different curriculum learning methods for training convolutional neural networks on the task of deformable pairwise 3D medical image registration.

28, TITLE: VisualGPT: Data-efficient Image Captioning By Balancing Visual Input and Linguistic Knowledge from Pretraining
AUTHORS: Jun Chen ; Han Guo ; Kai Yi ; Boyang Li ; Mohamed Elhoseiny
CATEGORY: cs.CV [cs.CV, cs.AI, cs.CL, cs.MM]
HIGHLIGHT: In this paper, we aim to improve the data efficiency of image captioning.

29, TITLE: Contrastive Self-supervised Neural Architecture Search
AUTHORS: Nam Nguyen ; J. Morris Chang
CATEGORY: cs.CV [cs.CV, cs.AI]
HIGHLIGHT: This paper proposes a novel cell-based neural architecture search algorithm (NAS), which completely alleviates the expensive costs of data labeling inherited from supervised learning.

30, TITLE: CSTR: A Classification Perspective on Scene Text Recognition
AUTHORS: Hongxiang Cai ; Jun Sun ; Yichao Xiong
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, we propose a new perspective on scene text recognition, in which we model the scene text recognition as an image classification problem.

31, TITLE: A Comprehensive Review of Computer-aided Whole-slide Image Analysis: from Datasets to Feature Extraction, Segmentation, Classification, and Detection Approaches
AUTHORS: CHEN LI et. al.
CATEGORY: cs.CV [cs.CV, cs.AI]
HIGHLIGHT: This paper reviews the methods of WSI analysis based on machine learning.

32, TITLE: Improving Action Quality Assessment Using ResNets and Weighted Aggregation
AUTHORS: SHAFKAT FARABI et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: To overcome this, we propose a learning-based weighted-averaging technique that can perform better.

33, TITLE: Efficient Two-Stream Network for Violence Detection Using Separable Convolutional LSTM
AUTHORS: Zahidul Islam ; Mohammad Rukonuzzaman ; Raiyan Ahmed ; Md. Hasanul Kabir ; Moshiur Farazi
CATEGORY: cs.CV [cs.CV, cs.LG]
HIGHLIGHT: In this work, we propose an efficient two-stream deep learning architecture leveraging Separable Convolutional LSTM (SepConvLSTM) and pre-trained MobileNet where one stream takes in background suppressed frames as inputs and other stream processes difference of adjacent frames.

34, TITLE: Do We Really Need Explicit Position Encodings for Vision Transformers?
AUTHORS: Xiangxiang Chu ; Bo Zhang ; Zhi Tian ; Xiaolin Wei ; Huaxia Xia
CATEGORY: cs.CV [cs.CV, cs.AI, cs.LG]
HIGHLIGHT: In this paper, we propose to employ a conditional position encoding scheme, which is conditioned on the local neighborhood of the input token.

35, TITLE: Learning Compositional Representation for Few-shot Visual Question Answering
AUTHORS: Dalu Guo ; Dacheng Tao
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: Therefore, in this paper, we propose to extract the attributes from the answers with enough data, which are later composed to constrain the learning of the few-shot ones. We generate the few-shot dataset of VQA with a variety of answers and their attributes without any human effort.

36, TITLE: MedAug: Contrastive Learning Leveraging Patient Metadata Improves Representations for Chest X-ray Interpretation
AUTHORS: YEN NHI TRUONG VU et. al.
CATEGORY: cs.CV [cs.CV, cs.LG, eess.IV]
HIGHLIGHT: In this work, we develop a method to select positive pairs coming from views of possibly different images through the use of patient metadata.

37, TITLE: A Deep Decomposition Network for Image Processing: A Case Study for Visible and Infrared Image Fusion
AUTHORS: Yu Fu ; Xiao-Jun Wu ; Josef Kittler
CATEGORY: cs.CV [cs.CV, eess.IV]
HIGHLIGHT: We propose a new image decomposition method based on convolutional neural network.

38, TITLE: Progressive Depth Learning for Single Image Dehazing
AUTHORS: YUDONG LIANG et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, a deep end-to-end model that iteratively estimates image depths and transmission maps is proposed to perform an effective depth prediction for hazy images and improve the dehazing performance with the guidance of depth information.

39, TITLE: Medical Transformer: Gated Axial-Attention for Medical Image Segmentation
AUTHORS: Jeya Maria Jose Valanarasu ; Poojan Oza ; Ilker Hacihaliloglu ; Vishal M. Patel
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: To this end, we propose a Gated Axial-Attention model which extends the existing architectures by introducing an additional control mechanism in the self-attention module.

40, TITLE: Training Custom Modality-specific U-Net Models with Weak Localizations for Improved Tuberculosis Segmentation and Localization
AUTHORS: Sivaramakrishnan Rajaraman ; Les Folio ; Jane Dimperio ; Philip Alderson ; Sameer Antani
CATEGORY: cs.CV [cs.CV, ACM-class: I.4, I.5]
HIGHLIGHT: In this study, we train custom chest X ray modality specific UNet models for semantic segmentation of Tuberculosis (TB) consistent findings.

41, TITLE: Do Generative Models Know Disentanglement? Contrastive Learning Is All You Need
AUTHORS: Xuanchi Ren ; Tao Yang ; Yuwang Wang ; Wenjun Zeng
CATEGORY: cs.CV [cs.CV, cs.AI, cs.LG]
HIGHLIGHT: We propose an unsupervised and model-agnostic method: Disentanglement via Contrast (DisCo) in the Variation Space.

42, TITLE: Rethinking Content and Style: Exploring Bias for Unsupervised Disentanglement
AUTHORS: Xuanchi Ren ; Tao Yang ; Yuwang Wang ; Wenjun Zeng
CATEGORY: cs.CV [cs.CV, cs.AI, cs.LG]
HIGHLIGHT: From the unsupervised disentanglement perspective, we rethink content and style and propose a formulation for unsupervised C-S disentanglement based on our assumption that different factors are of different importance and popularity for image reconstruction, which serves as a data bias.

43, TITLE: Deepfake Video Detection Using Convolutional Vision Transformer
AUTHORS: Deressa Wodajo ; Solomon Atnafu
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this work, we propose a Convolutional Vision Transformer for the detection of Deepfakes.

44, TITLE: Image Captioning Using Deep Stacked LSTMs, Contextual Word Embeddings and Data Augmentation
AUTHORS: Sulabh Katiyar ; Samir Kumar Borgohain
CATEGORY: cs.CV [cs.CV, cs.AI, cs.LG, cs.MM, cs.NE]
HIGHLIGHT: We propose to use Inception-ResNet Convolutional Neural Network as encoder to extract features from images, Hierarchical Context based Word Embeddings for word representations and a Deep Stacked Long Short Term Memory network as decoder, in addition to using Image Data Augmentation to avoid over-fitting.

45, TITLE: Wider Vision: Enriching Convolutional Neural Networks Via Alignment to External Knowledge Bases
AUTHORS: Xuehao Liu ; Sarah Jane Delany ; Susan McKeever
CATEGORY: cs.CV [cs.CV, cs.AI]
HIGHLIGHT: The aim of our work is to explain and expand CNNs models via the mirroring or alignment of CNN to an external knowledge base.

46, TITLE: Hierarchical Cell-to-Tissue Graph Representations for Breast Cancer Subtyping in Digital Pathology
AUTHORS: PUSHPAK PATI et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: We propose a novel hierarchical entity-graph representation to depict a tissue specimen, which encodes multiple pathologically relevant entity types, intra- and inter-level entity-to-entity interactions.

47, TITLE: SALT: A Semi-automatic Labeling Tool for RGB-D Video Sequences
AUTHORS: Dennis Stumpf ; Stephan Krau� ; Gerd Reis ; Oliver Wasenm�ller ; Didier Stricker
CATEGORY: cs.CV [cs.CV, cs.LG, cs.RO]
HIGHLIGHT: In this paper, we introduce SALT, a tool to semi-automatically annotate RGB-D video sequences to generate 3D bounding boxes for full six Degrees of Freedom (DoF) object poses, as well as pixel-level instance segmentation masks for both RGB and depth.

48, TITLE: Subspace-Based Feature Fusion From Hyperspectral And Multispectral Image For Land Cover Classification
AUTHORS: Juan Ram�rez ; H�ctor Vargas ; Jos� Ignacio Mart�nez ; Henry Arguello
CATEGORY: cs.CV [cs.CV, eess.IV]
HIGHLIGHT: In this paper, a feature fusion method from HS and MS images for pixel-based classification is proposed.

49, TITLE: Self-Supervised Learning Via Multi-Transformation Classification for Action Recognition
AUTHORS: Duc Quang Vu ; Ngan T. H. Le ; Jia-Ching Wang
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, we introduce a self-supervised video representation learning method based on the multi-transformation classification to efficiently classify human actions.

50, TITLE: Generator Surgery for Compressed Sensing
AUTHORS: NIKLAS SMEDEMARK-MARGULIES et. al.
CATEGORY: cs.CV [cs.CV, eess.IV]
HIGHLIGHT: We introduce a method for achieving low representation error using generators as signal priors.

51, TITLE: CellTrack R-CNN: A Novel End-To-End Deep Neural Network for Cell Segmentation and Tracking in Microscopy Images
AUTHORS: YUQIAN CHEN et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this study, we propose a novel approach to combine cell segmentation and cell tracking into a unified end-to-end deep learning based framework, where cell detection and segmentation are performed with a current instance segmentation pipeline and cell tracking is implemented by integrating Siamese Network with the pipeline.

52, TITLE: Phase Space Reconstruction Network for Lane Intrusion Action Recognition
AUTHORS: Ruiwen Zhang ; Zhidong Deng ; Hongsen Lin ; Hongchao Lu
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, we propose a novel object-level phase space reconstruction network (PSRNet) for motion time series classification, aiming to recognize lane intrusion actions that occur 150m ahead through a monocular camera fixed on moving vehicle.

53, TITLE: Style and Pose Control for Image Synthesis of Humans from A Single Monocular View
AUTHORS: Kripasindhu Sarkar ; Vladislav Golyanik ; Lingjie Liu ; Christian Theobalt
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: We, therefore, propose a new method for synthesising photo-realistic human images with explicit control over pose and part-based appearance, i.e., StylePoseGAN, where we extend a non-controllable generator to accept conditioning of pose and appearance separately.

54, TITLE: Adversarial Shape Learning for Building Extraction in VHR Remote Sensing Images
AUTHORS: Lei Ding ; Hao Tang ; Yahui Liu ; Yilei Shi ; Lorenzo Bruzzone
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In the proposed ASLNet, we introduce the adversarial learning strategy to explicitly model the shape constraints, as well as a CNN shape regularizer to strengthen the embedding of shape features.

55, TITLE: On Interaction Between Augmentations and Corruptions in Natural Corruption Robustness
AUTHORS: Eric Mintun ; Alexander Kirillov ; Saining Xie
CATEGORY: cs.CV [cs.CV, cs.LG]
HIGHLIGHT: On Interaction Between Augmentations and Corruptions in Natural Corruption Robustness

56, TITLE: Decoupled and Memory-Reinforced Networks: Towards Effective Feature Learning for One-Step Person Search
AUTHORS: Chuchu Han ; Zhedong Zheng ; Changxin Gao ; Nong Sang ; Yi Yang
CATEGORY: cs.CV [cs.CV, cs.AI]
HIGHLIGHT: To overcome these problems, we propose a decoupled and memory-reinforced network (DMRNet).

57, TITLE: Person Re-identification Based on Robust Features in Open-world
AUTHORS: Yaguan Qian ; Anlin Sun
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, we propose a low-cost and high-efficiency method to solve shortcomings of the existing re-ID research, such as unreliable feature selection, low efficiency of feature extraction, single research variable, etc. At the same time, to verify the effectiveness of our method, we provide a miniature dataset which is closer to the real world and includes pedestrian changing clothes and cross-modality factor variables fusion.

58, TITLE: EMDS-5: Environmental Microorganism Image Dataset Fifth Version for Multiple Image Analysis Tasks
AUTHORS: ZIHAN LI et. al.
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: Six methods are used for single-object image segmentation, while k-means and U-net are used for multi-object segmentation.We extract nine features from the images in EMDS-5 and use the Support Vector Machine classifier for testing.

59, TITLE: Exploring Knowledge Distillation of A Deep Neural Network for Multi-Script Identification
AUTHORS: Shuvayan Ghosh Dastidar ; Kalpita Dutta ; Nibaran Das ; Mahantapas Kundu ; Mita Nasipuri
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, we explore dark knowledge transfer approach using long short-term memory(LSTM) and CNN based assistant model and various deep neural networks as the teacher model, with a simple CNN based student network, in this domain of multi-script identification from natural scene text images.

60, TITLE: Post-hoc Overall Survival Time Prediction from Brain MRI
AUTHORS: Renato Hermoza ; Gabriel Maicas ; Jacinto C. Nascimento ; Gustavo Carneiro
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this paper, we introduce a new post-hoc method for OS time prediction that does not require segmentation map annotation for training.

61, TITLE: Analyzing Overfitting Under Class Imbalance in Neural Networks for Image Segmentation
AUTHORS: Zeju Li ; Konstantinos Kamnitsas ; Ben Glocker
CATEGORY: cs.CV [cs.CV]
HIGHLIGHT: In this study, we provide new insights on the problem of overfitting under class imbalance by inspecting the network behavior.

62, TITLE: WaNet -- Imperceptible Warping-based Backdoor Attack
AUTHORS: Anh Nguyen ; Anh Tran
CATEGORY: cs.CR [cs.CR, cs.CV]
HIGHLIGHT: In this paper, we instead propose using warping-based triggers.

63, TITLE: Differentially Private Supervised Manifold Learning with Applications Like Private Image Retrieval
AUTHORS: Praneeth Vepakomma ; Julia Balla ; Ramesh Raskar
CATEGORY: cs.LG [cs.LG, cs.CR, cs.CV, cs.DB, cs.DC]
HIGHLIGHT: 1) We present a novel differentially private method \textit{PrivateMail} for supervised manifold learning, the first of its kind to our knowledge.

64, TITLE: The Effects of Image Distribution and Task on Adversarial Robustness
AUTHORS: Owen Kunhardt ; Arturo Deza ; Tomaso Poggio
CATEGORY: cs.LG [cs.LG, cs.CV]
HIGHLIGHT: In this paper, we propose an adaptation to the area under the curve (AUC) metric to measure the adversarial robustness of a model over a particular $\epsilon$-interval $[\epsilon_0, \epsilon_1]$ (interval of adversarial perturbation strengths) that facilitates unbiased comparisons across models when they have different initial $\epsilon_0$ performance.

65, TITLE: Learning Purified Feature Representations from Task-irrelevant Labels
AUTHORS: YINGHUI LI et. al.
CATEGORY: cs.LG [cs.LG, cs.CV]
HIGHLIGHT: In this paper, we propose a novel learning framework called PurifiedLearning to exploit task-irrelevant features extracted from task-irrelevant labels when training models on small-scale datasets.

66, TITLE: Probing Multimodal Embeddings for Linguistic Properties: The Visual-Semantic Case
AUTHORS: Adam Dahlgren Lindstr�m ; Suna Bensch ; Johanna Bj�rklund ; Frank Drewes
CATEGORY: cs.LG [cs.LG, cs.CL, cs.CV]
HIGHLIGHT: Semantic embeddings have advanced the state of the art for countless natural language processing tasks, and various extensions to multimodal domains, such as visual-semantic embeddings, have been proposed.

67, TITLE: GroupifyVAE: from Group-based Definition to VAE-based Unsupervised Representation Disentanglement
AUTHORS: TAO YANG et. al.
CATEGORY: cs.LG [cs.LG, cs.CV]
HIGHLIGHT: In this paper, we address VAE-based unsupervised disentanglement by leveraging the constraints derived from the Group Theory based definition as the non-probabilistic inductive bias.

68, TITLE: Weak NAS Predictors Are All You Need
AUTHORS: JUNRU WU et. al.
CATEGORY: cs.LG [cs.LG, cs.CV, stat.ML]
HIGHLIGHT: In this paper, we shift the paradigm from finding a complicated predictor that covers the whole architecture space to a set of weaker predictors that progressively move towards the high-performance sub-space.

69, TITLE: Learning Neural Network Subspaces
AUTHORS: Mitchell Wortsman ; Maxwell Horton ; Carlos Guestrin ; Ali Farhadi ; Mohammad Rastegari
CATEGORY: cs.LG [cs.LG, cs.CV]
HIGHLIGHT: In contrast we aim to leverage both property (1) and (2) with a single method and in a single training run.

70, TITLE: On Fast Adversarial Robustness Adaptation in Model-Agnostic Meta-Learning
AUTHORS: REN WANG et. al.
CATEGORY: cs.LG [cs.LG, cs.AI, cs.CV]
HIGHLIGHT: We propose a general but easily-optimized robustness-regularized meta-learning framework, which allows the use of unlabeled data augmentation, fast adversarial attack generation, and computationally-light fine-tuning.

71, TITLE: BSQ: Exploring Bit-Level Sparsity for Mixed-Precision Neural Network Quantization
AUTHORS: Huanrui Yang ; Lin Duan ; Yiran Chen ; Hai Li
CATEGORY: cs.LG [cs.LG, cs.CV]
HIGHLIGHT: This work proposes bit-level sparsity quantization (BSQ) to tackle the mixed-precision quantization from a new angle of inducing bit-level sparsity.

72, TITLE: SSFG: Stochastically Scaling Features and Gradients for Regularizing Graph Convolution Networks
AUTHORS: Haimin Zhang ; Min Xu
CATEGORY: cs.LG [cs.LG, cs.CV]
HIGHLIGHT: In this paper, we present a stochastic regularization method to address this issue.

73, TITLE: Combining Spiking Neural Network and Artificial Neural Network for Enhanced Image Classification
AUTHORS: Naoya Muramatsu ; Hai-Tao Yu
CATEGORY: cs.NE [cs.NE, cs.CV, cs.LG, 68T05 (Primary) 68T05 (Secondary), I.2.6]
HIGHLIGHT: To this end, we combine an ANN and an SNN to build versatile hybrid neural networks (HNNs) that improve the concerned performance.

74, TITLE: 3D Vision-guided Pick-and-Place Using Kuka LBR Iiwa Robot
AUTHORS: Hanlin Niu ; Ze Ji ; Zihang Zhu ; Hujun Yin ; Joaquin Carrasco
CATEGORY: cs.RO [cs.RO, cs.CV]
HIGHLIGHT: This paper presents the development of a control system for vision-guided pick-and-place tasks using a robot arm equipped with a 3D camera.

75, TITLE: Classification of COVID-19 Via Homology of CT-SCAN
AUTHORS: Sohail Iqbal ; H. Fareed Ahmed ; Talha Qaiser ; Muhammad Imran Qureshi ; Nasir Rajpoot
CATEGORY: eess.IV [eess.IV, cs.CV, math.AT]
HIGHLIGHT: In this article, We propose a novel approach to detect SARS-CoV-2 using CT-scan images.

76, TITLE: Tchebichef Transform Domain-based Deep Learning Architecture for Image Super-resolution
AUTHORS: Ahlad Kumar ; Harsh Vardhan Singh
CATEGORY: eess.IV [eess.IV, cs.CV, cs.LG]
HIGHLIGHT: In this paper, we propose a deep learning based image super-resolution architecture in Tchebichef transform domain.

77, TITLE: RCoNet: Deformable Mutual Information Maximization and High-order Uncertainty-aware Learning for Robust COVID-19 Detection
AUTHORS: Shunjie Dong ; Qianqian Yang ; Yu Fu ; Mei Tian ; Cheng Zhuo
CATEGORY: eess.IV [eess.IV, cs.CV]
HIGHLIGHT: To alleviate these concerns, we propose a novel deep network named {\em RCoNet$^k_s$} for robust COVID-19 detection which employs {\em Deformable Mutual Information Maximization} (DeIM), {\em Mixed High-order Moment Feature} (MHMF) and {\em Multi-expert Uncertainty-aware Learning} (MUL).