Improved Regularization of Convolutional Neural Networks with Cutout

Abstract

Convolutional neural networks are capable of learning powerful representational spaces, which are necessary for tackling complex learning tasks. However, due to the model capacity required to capture such representations, they are often susceptible to overfitting and therefore require proper regularization in order to generalize well. In this paper, we show that the simple regularization technique of randomly masking out square regions of input during training, which we call cutout, can be used to improve the robustness and overall performance of convolutional neural networks. Not only is this method extremely easy to implement, but we also demonstrate that it can be used in conjunction with existing forms of data augmentation and other regularizers to further improve model performance. We evaluate this method by applying it to current state-of-the-art architectures on the CIFAR-10, CIFAR-100, and SVHN datasets, yielding new state-of-the-art results of 2.56%, 15.20%, and 1.30% test error respectively. Code is available at https://github.com/uoguelph-mlrg/Cutout

References in corpus (7)

• Improving neural networks by preventing co-adaptation of feature detectors
• DeepPose: Human Pose Estimation via Deep Neural Networks
• An Analysis of Deep Neural Network Models for Practical Applications
• Smart Augmentation - Learning an Optimal Data Augmentation Strategy
• Deep Image: Scaling up Image Recognition
• Towards Dropout Training for Convolutional Neural Networks
• Shake-Shake regularization

Cited by in corpus (678)

• YOLOv4: Optimal Speed and Accuracy of Object Detection
• A Simple Framework for Contrastive Learning of Visual Representations
• SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition
• FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence
• Neural Architecture Search: A Survey
• AutoML: A Survey of the State-of-the-Art
• Unsupervised Data Augmentation for Consistency Training
• Domain Generalization: A Survey
• Training Generative Adversarial Networks with Limited Data
• A Survey on Deep Semi-supervised Learning
• Automatically designing CNN architectures using genetic algorithm for image classification
• AutoAugment: Learning Augmentation Policies from Data
• A Survey on Evolutionary Neural Architecture Search
• A Comprehensive Survey of Image Augmentation Techniques for Deep Learning
• CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features
• MixMatch: A Holistic Approach to Semi-Supervised Learning
• AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty
• Data Augmentation using Random Image Cropping and Patching for Deep CNNs
• Hyper-Parameter Optimization: A Review of Algorithms and Applications
• Born Again Neural Networks
• PC-DARTS: Partial Channel Connections for Memory-Efficient Architecture Search
• Avoiding Overfitting: A Survey on Regularization Methods for Convolutional Neural Networks
• Do ImageNet Classifiers Generalize to ImageNet?
• Lookahead Optimizer: k steps forward, 1 step back
• Learning Confidence for Out-of-Distribution Detection in Neural Networks
• A Simple Semi-Supervised Learning Framework for Object Detection
• Generalized Out-of-Distribution Detection: A Survey
• R-Drop: Regularized Dropout for Neural Networks
• A 64-core mixed-signal in-memory compute chip based on phase-change memory for deep neural network inference
• ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware
• SNAS: Stochastic Neural Architecture Search
• CSI: Novelty Detection via Contrastive Learning on Distributionally Shifted Instances
• Understanding and Robustifying Differentiable Architecture Search
• GPipe: Efficient Training of Giant Neural Networks using Pipeline Parallelism
• Random Search and Reproducibility for Neural Architecture Search
• Omni-frequency Channel-selection Representations for Unsupervised Anomaly Detection
• GridMask Data Augmentation
• Single Model Deep Learning on Imbalanced Small Datasets for Skin Lesion Classification
• A survey on Semi-, Self- and Unsupervised Learning for Image Classification
• Real-Time Apple Detection System Using Embedded Systems With Hardware Accelerators: An Edge AI Application
• One-Shot Neural Architecture Search via Self-Evaluated Template Network
• A Survey on Neural Architecture Search
• Adversarial Weight Perturbation Helps Robust Generalization
• PseudoSeg: Designing Pseudo Labels for Semantic Segmentation
• Realistic Evaluation of Deep Semi-Supervised Learning Algorithms
• A Comprehensive Survey of Neural Architecture Search: Challenges and Solutions
• Mixed Precision Quantization of ConvNets via Differentiable Neural Architecture Search
• Quantifying Generalization in Reinforcement Learning
• A Survey of Zero-shot Generalisation in Deep Reinforcement Learning
• ShakeDrop Regularization for Deep Residual Learning
• SPICE: Semantic Pseudo-labeling for Image Clustering
• Natural Adversarial Examples
• Measuring Robustness to Natural Distribution Shifts in Image Classification
• Progressive Neural Architecture Search
• Evaluating the Search Phase of Neural Architecture Search
• Robust and generalizable embryo selection based on artificial intelligence and time-lapse image sequences
• Semi-supervised semantic segmentation needs strong, varied perturbations
• Twin Contrastive Learning for Online Clustering
• Automated data processing and feature engineering for deep learning and big data applications: a survey
• Path-Level Network Transformation for Efficient Architecture Search
• Uncovering the Limits of Adversarial Training against Norm-Bounded Adversarial Examples
• Neural Attention Distillation: Erasing Backdoor Triggers from Deep Neural Networks
• How Deep Learning Sees the World: A Survey on Adversarial Attacks & Defenses
• Progressive Differentiable Architecture Search: Bridging the Depth Gap between Search and Evaluation
• Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels
• Deep Convolutions for In-Depth Automated Rock Typing
• A New Dataset, Poisson GAN and AquaNet for Underwater Object Grabbing
• Further advantages of data augmentation on convolutional neural networks
• FastReID: A Pytorch Toolbox for General Instance Re-identification
• The Origins and Prevalence of Texture Bias in Convolutional Neural Networks
• Query2Label: A Simple Transformer Way to Multi-Label Classification
• BDANet: Multiscale Convolutional Neural Network with Cross-directional Attention for Building Damage Assessment from Satellite Images
• Image Augmentations for GAN Training
• A Comprehensive Study of Deep Video Action Recognition
• Graph HyperNetworks for Neural Architecture Search
• Cross-Domain Few-Shot Classification via Learned Feature-Wise Transformation
• Recent Advances in Object Detection in the Age of Deep Convolutional Neural Networks
• PP-OCR: A Practical Ultra Lightweight OCR System
• Fixup Initialization: Residual Learning Without Normalization
• NAS evaluation is frustratingly hard
• Streaming convolutional neural networks for end-to-end learning with multi-megapixel images
• Second-order Non-local Attention Networks for Person Re-identification
• Fixing Data Augmentation to Improve Adversarial Robustness
• Sharpness-Aware Minimization for Efficiently Improving Generalization
• FMix: Enhancing Mixed Sample Data Augmentation
• See Better Before Looking Closer: Weakly Supervised Data Augmentation Network for Fine-Grained Visual Classification
• Pose-Aware Instance Segmentation Framework from Cone Beam CT Images for Tooth Segmentation
• Differentiable Augmentation for Data-Efficient GAN Training
• Label-Only Membership Inference Attacks
• Data augmentation instead of explicit regularization
• SaliencyMix: A Saliency Guided Data Augmentation Strategy for Better Regularization
• Improving Robustness Without Sacrificing Accuracy with Patch Gaussian Augmentation
• Incomplete Descriptor Mining with Elastic Loss for Person Re-Identification
• Best Practices for Scientific Research on Neural Architecture Search
• D-VAE: A Variational Autoencoder for Directed Acyclic Graphs
• Towards Automated Deep Learning: Efficient Joint Neural Architecture and Hyperparameter Search
• An overview of mixing augmentation methods and augmentation strategies
• Stabilizing Differentiable Architecture Search via Perturbation-based Regularization
• NAS-Bench-1Shot1: Benchmarking and Dissecting One-shot Neural Architecture Search
• Putting An End to End-to-End: Gradient-Isolated Learning of Representations
• Occlusion Handling in Generic Object Detection: A Review
• Subject-Aware Contrastive Learning for Biosignals
• Object Detection Using Sim2Real Domain Randomization for Robotic Applications
• In Defense of Pseudo-Labeling: An Uncertainty-Aware Pseudo-label Selection Framework for Semi-Supervised Learning
• NSGA-Net: Neural Architecture Search using Multi-Objective Genetic Algorithm
• Reference-guided Pseudo-Label Generation for Medical Semantic Segmentation
• Unlabeled Data Improves Adversarial Robustness
• BNAS:An Efficient Neural Architecture Search Approach Using Broad Scalable Architecture
• Bridging the Gap between Sample-based and One-shot Neural Architecture Search with BONAS
• A Good Image Generator Is What You Need for High-Resolution Video Synthesis
• Efficient Multi-objective Neural Architecture Search via Lamarckian Evolution
• EEEA-Net: An Early Exit Evolutionary Neural Architecture Search
• Maximum-Entropy Adversarial Data Augmentation for Improved Generalization and Robustness
• ASAP: Architecture Search, Anneal and Prune
• Asymmetric Loss For Multi-Label Classification
• Bayesian Meta-Learning for the Few-Shot Setting via Deep Kernels
• A Systematic Survey of Regularization and Normalization in GANs
• Self-Supervised Learning with Data Augmentations Provably Isolates Content from Style
• You Only Train Once: Learning a General Anomaly Enhancement Network with Random Masks for Hyperspectral Anomaly Detection
• EDropout: Energy-Based Dropout and Pruning of Deep Neural Networks
• A Survey of Deep Learning for Scientific Discovery
• A Bayesian Federated Learning Framework with Online Laplace Approximation
• Remix: Rebalanced Mixup
• Data Augmentation for Object Detection via Progressive and Selective Instance-Switching
• Taming Unbalanced Training Workloads in Deep Learning with Partial Collective Operations
• Domain generalization in deep learning-based mass detection in mammography: A large-scale multi-center study
• Effective and Efficient Dropout for Deep Convolutional Neural Networks
• i-Mix: A Domain-Agnostic Strategy for Contrastive Representation Learning
• Dataset Condensation with Differentiable Siamese Augmentation
• Affinity and Diversity: Quantifying Mechanisms of Data Augmentation
• Multi-Sample Dropout for Accelerated Training and Better Generalization
• Automatic Data Augmentation for Generalization in Deep Reinforcement Learning
• Augment your batch: better training with larger batches
• How Robust is 3D Human Pose Estimation to Occlusion?
• MEAL V2: Boosting Vanilla ResNet-50 to 80%+ Top-1 Accuracy on ImageNet without Tricks
• Anatomy-XNet: An Anatomy Aware Convolutional Neural Network for Thoracic Disease Classification in Chest X-rays
• Generative Teaching Networks: Accelerating Neural Architecture Search by Learning to Generate Synthetic Training Data
• Adversarial Examples Improve Image Recognition
• VideoMix: Rethinking Data Augmentation for Video Classification
• Physical Knowledge Enhanced Deep Neural Network for Sea Surface Temperature Prediction
• Network Randomization: A Simple Technique for Generalization in Deep Reinforcement Learning
• Survey of Dropout Methods for Deep Neural Networks
• Self-Augmentation: Generalizing Deep Networks to Unseen Classes for Few-Shot Learning
• Consistency Regularization for Generative Adversarial Networks
• Identifying Melanoma Images using EfficientNet Ensemble: Winning Solution to the SIIM-ISIC Melanoma Classification Challenge
• Unlearnable Examples: Making Personal Data Unexploitable
• Graph Convolutional Neural Networks with Node Transition Probability-based Message Passing and DropNode Regularization
• Conservative-Progressive Collaborative Learning for Semi-supervised Semantic Segmentation
• Overfitting in adversarially robust deep learning
• Model Patching: Closing the Subgroup Performance Gap with Data Augmentation
• Stabilizing DARTS with Amended Gradient Estimation on Architectural Parameters
• Deep Weakly-Supervised Learning Methods for Classification and Localization in Histology Images: A Survey
• Scale-Equivariant Steerable Networks
• Recent Advances in Adversarial Training for Adversarial Robustness
• Position, Padding and Predictions: A Deeper Look at Position Information in CNNs
• Whitening for Self-Supervised Representation Learning
• Fingerprint Presentation Attack Detection by Channel-wise Feature Denoising
• Learning perturbation sets for robust machine learning
• ABC: Auxiliary Balanced Classifier for Class-imbalanced Semi-supervised Learning
• Domain Generalization with MixStyle
• You Only Search Once: Single Shot Neural Architecture Search via Direct Sparse Optimization
• DADA: Differentiable Automatic Data Augmentation
• Convolution with even-sized kernels and symmetric padding
• ResizeMix: Mixing Data with Preserved Object Information and True Labels
• Learning Loss for Test-Time Augmentation
• Synthetic Occlusion Augmentation with Volumetric Heatmaps for the 2018 ECCV PoseTrack Challenge on 3D Human Pose Estimation
• Ensemble learning in CNN augmented with fully connected subnetworks
• Fingerprint Presentation Attack Detector Using Global-Local Model
• Contrastive Self-supervised Neural Architecture Search
• An Empirical Evaluation on Robustness and Uncertainty of Regularization Methods
• Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentation
• Accelerating Deep Learning by Focusing on the Biggest Losers
• Cut-Thumbnail: A Novel Data Augmentation for Convolutional Neural Network
• Data Augmentation Revisited: Rethinking the Distribution Gap between Clean and Augmented Data
• Assume, Augment and Learn: Unsupervised Few-Shot Meta-Learning via Random Labels and Data Augmentation
• sharpDARTS: Faster and More Accurate Differentiable Architecture Search
• Learning Contrastive Self-Distillation for Ultra-Fine-Grained Visual Categorization Targeting Limited Samples
• REB: Reducing Biases in Representation for Industrial Anomaly Detection
• Deep Gamblers: Learning to Abstain with Portfolio Theory
• DropAttention: A Regularization Method for Fully-Connected Self-Attention Networks
• MaxUp: A Simple Way to Improve Generalization of Neural Network Training
• An original framework for Wheat Head Detection using Deep, Semi-supervised and Ensemble Learning within Global Wheat Head Detection (GWHD) Dataset
• PatchUp: A Feature-Space Block-Level Regularization Technique for Convolutional Neural Networks
• CReST: A Class-Rebalancing Self-Training Framework for Imbalanced Semi-Supervised Learning
• Deeply Self-Supervised Contour Embedded Neural Network Applied to Liver Segmentation
• Reinforced Evolutionary Neural Architecture Search
• VarifocalNet: An IoU-aware Dense Object Detector
• Towards Better Accuracy-efficiency Trade-offs: Divide and Co-training
• QuartzNet: Deep Automatic Speech Recognition with 1D Time-Channel Separable Convolutions
• Towards Lossless ANN-SNN Conversion under Ultra-Low Latency with Dual-Phase Optimization
• PConv: Simple yet Effective Convolutional Layer for Generative Adversarial Network
• Regularizing Class-wise Predictions via Self-knowledge Distillation
• Optimized Generic Feature Learning for Few-shot Classification across Domains
• GOLD-NAS: Gradual, One-Level, Differentiable
• I2F: A Unified Image-to-Feature Approach for Domain Adaptive Semantic Segmentation
• Meta-Learned Confidence for Few-shot Learning
• Improving Auto-Augment via Augmentation-Wise Weight Sharing
• Unbiased Teacher for Semi-Supervised Object Detection
• Parting with Illusions about Deep Active Learning
• CoDA: Contrast-enhanced and Diversity-promoting Data Augmentation for Natural Language Understanding
• QMaxViT-Unet+: A Query-Based MaxViT-Unet with Edge Enhancement for Scribble-Supervised Segmentation of Medical Images
• Entropic gradient descent algorithms and wide flat minima
• Self-Supervised Relational Reasoning for Representation Learning
• Assessing Post-Disaster Damage from Satellite Imagery using Semi-Supervised Learning Techniques
• On Feature Normalization and Data Augmentation
• Latency-Aware Differentiable Neural Architecture Search
• Instant-Teaching: An End-to-End Semi-Supervised Object Detection Framework
• Learning Invariances in Neural Networks
• Data-driven Crop Growth Simulation on Time-varying Generated Images using Multi-conditional Generative Adversarial Networks
• Exploring Data Augmentation for Multi-Modality 3D Object Detection
• Improved Consistency Regularization for GANs
• RC-DARTS: Resource Constrained Differentiable Architecture Search
• Learning Implicitly Recurrent CNNs Through Parameter Sharing
• Theory-Inspired Path-Regularized Differential Network Architecture Search
• GradAug: A New Regularization Method for Deep Neural Networks
• Equalization Loss for Long-Tailed Object Recognition
• Lipschitz regularized Deep Neural Networks generalize and are adversarially robust
• Why Does Little Robustness Help? A Further Step Towards Understanding Adversarial Transferability
• Batch DropBlock Network for Person Re-identification and Beyond
• Self-Supervised Time Series Representation Learning by Inter-Intra Relational Reasoning
• Regularizing Deep Networks with Semantic Data Augmentation
• Poisoning the Unlabeled Dataset of Semi-Supervised Learning
• Generalized Jensen-Shannon Divergence Loss for Learning with Noisy Labels
• CIFAR10 to Compare Visual Recognition Performance between Deep Neural Networks and Humans
• XNAS: Neural Architecture Search with Expert Advice
• Improved robustness to adversarial examples using Lipschitz regularization of the loss
• Boundary thickness and robustness in learning models
• Unsupervised Semantic Aggregation and Deformable Template Matching for Semi-Supervised Learning
• GraphCrop: Subgraph Cropping for Graph Classification
• ST++: Make Self-training Work Better for Semi-supervised Semantic Segmentation
• An Empirical Survey of Data Augmentation for Limited Data Learning in NLP
• Adaptive Betweenness Clustering for Semi-Supervised Domain Adaptation
• AIM 2020 Challenge on Real Image Super-Resolution: Methods and Results
• GuidedMixup: An Efficient Mixup Strategy Guided by Saliency Maps
• Progressive DARTS: Bridging the Optimization Gap for NAS in the Wild
• A Squeeze-and-Excitation and Transformer based Cross-task System for Environmental Sound Recognition
• A Unified Survey on Anomaly, Novelty, Open-Set, and Out-of-Distribution Detection: Solutions and Future Challenges
• Multi-Label Knowledge Distillation
• ExtremeC3Net: Extreme Lightweight Portrait Segmentation Networks using Advanced C3-modules
• PatchCleanser: Certifiably Robust Defense against Adversarial Patches for Any Image Classifier
• Joint Neural Architecture Search and Quantization
• ASAM: Adaptive Sharpness-Aware Minimization for Scale-Invariant Learning of Deep Neural Networks
• CutDepth:Edge-aware Data Augmentation in Depth Estimation
• HSMix: Hard and Soft Mixing Data Augmentation for Medical Image Segmentation
• BlockQNN: Efficient Block-wise Neural Network Architecture Generation
• On the Convergence of AdaBound and its Connection to SGD
• Rainbow Memory: Continual Learning with a Memory of Diverse Samples
• DP-InstaHide: Provably Defusing Poisoning and Backdoor Attacks with Differentially Private Data Augmentations
• On the effectiveness of adversarial training against common corruptions
• Image Based Identification of Ghanaian Timbers Using the XyloTron: Opportunities, Risks and Challenges
• TDAPNet: Prototype Network with Recurrent Top-Down Attention for Robust Object Classification under Partial Occlusion
• PseudoAugment: Learning to Use Unlabeled Data for Data Augmentation in Point Clouds
• Neural Architecture Generator Optimization
• Wandering Within a World: Online Contextualized Few-Shot Learning
• The Deep Bootstrap Framework: Good Online Learners are Good Offline Generalizers
• Active Testing: Sample-Efficient Model Evaluation
• Sparse and Dense Data with CNNs: Depth Completion and Semantic Segmentation
• Multi-Objective Evolutionary Design of Deep Convolutional Neural Networks for Image Classification
• Self-Knowledge Distillation with Progressive Refinement of Targets
• Artificial and beneficial -- Exploiting artificial images for aerial vehicle detection
• Self-supervised Pre-training with Hard Examples Improves Visual Representations
• MatchboxNet: 1D Time-Channel Separable Convolutional Neural Network Architecture for Speech Commands Recognition
• RoIMix: Proposal-Fusion among Multiple Images for Underwater Object Detection
• MixStyle Neural Networks for Domain Generalization and Adaptation
• Transfer Learning Gaussian Anomaly Detection by Fine-tuning Representations
• Stochastic Weight Averaging in Parallel: Large-Batch Training that Generalizes Well
• Multinomial Distribution Learning for Effective Neural Architecture Search
• One-Shot Neural Architecture Search via Compressive Sensing
• Temporal Complementary Learning for Video Person Re-Identification
• EnAET: A Self-Trained framework for Semi-Supervised and Supervised Learning with Ensemble Transformations
• Revisiting 3D ResNets for Video Recognition
• A Large Multi-Target Dataset of Common Bengali Handwritten Graphemes
• Rafiki: Machine Learning as an Analytics Service System
• AT-ST: Self-Training Adaptation Strategy for OCR in Domains with Limited Transcriptions
• Anatomically aware dual-hop learning for pulmonary embolism detection in CT pulmonary angiograms
• IFQA: Interpretable Face Quality Assessment
• Semi-Supervised Domain Generalization with Stochastic StyleMatch
• Auxiliary Task Reweighting for Minimum-data Learning
• Core Box Image Recognition and its Improvement with a New Augmentation Technique
• Saliency is a Possible Red Herring When Diagnosing Poor Generalization
• BdSL36: A Dataset for Bangladeshi Sign Letters Recognition
• Generalization in Machine Learning via Analytical Learning Theory
• ProgFed: Effective, Communication, and Computation Efficient Federated Learning by Progressive Training
• Fine-Grained Neural Architecture Search
• Interpolation-based semi-supervised learning for object detection
• Robust Image Classification Using A Low-Pass Activation Function and DCT Augmentation
• LocalDrop: A Hybrid Regularization for Deep Neural Networks
• Dual-Cross Central Difference Network for Face Anti-Spoofing
• An Empirical Analysis of the Impact of Data Augmentation on Knowledge Distillation
• Self-supervised Temporal Discriminative Learning for Video Representation Learning
• BlockSwap: Fisher-guided Block Substitution for Network Compression on a Budget
• KeepAugment: A Simple Information-Preserving Data Augmentation Approach
• LibFewShot: A Comprehensive Library for Few-shot Learning
• StackMix and Blot Augmentations for Handwritten Text Recognition
• Monocular UAV Localisation with Deep Learning and Uncertainty Propagation
• Robust Asymmetric Heterogeneous Federated Learning with Corrupted Clients
• AID: Pushing the Performance Boundary of Human Pose Estimation with Information Dropping Augmentation
• RealMix: Towards Realistic Semi-Supervised Deep Learning Algorithms
• Robust Mutual Learning for Semi-supervised Semantic Segmentation
• Improving 3D Object Detection through Progressive Population Based Augmentation
• ContextMix: A context-aware data augmentation method for industrial visual inspection systems
• SwitchOut: an Efficient Data Augmentation Algorithm for Neural Machine Translation
• Comparing to Learn: Surpassing ImageNet Pretraining on Radiographs By Comparing Image Representations
• Neural Network Pruning with Residual-Connections and Limited-Data
• Federated Semi-Supervised Learning for COVID Region Segmentation in Chest CT using Multi-National Data from China, Italy, Japan
• Scaling-Translation-Equivariant Networks with Decomposed Convolutional Filters
• Robustness-Reinforced Knowledge Distillation with Correlation Distance and Network Pruning
• DVOLVER: Efficient Pareto-Optimal Neural Network Architecture Search
• SpecMix : A Mixed Sample Data Augmentation method for Training withTime-Frequency Domain Features
• Learning Optimal Conformal Classifiers
• Distilling Localization for Self-Supervised Representation Learning
• FenceMask: A Data Augmentation Approach for Pre-extracted Image Features
• Towards Good Practices in Self-supervised Representation Learning
• How Useful is Self-Supervised Pretraining for Visual Tasks?
• Cyclic Differentiable Architecture Search
• GradInit: Learning to Initialize Neural Networks for Stable and Efficient Training
• Efficient Neural Architecture Search via Proximal Iterations
• 3DIoUMatch: Leveraging IoU Prediction for Semi-Supervised 3D Object Detection
• Representative Forgery Mining for Fake Face Detection
• Generalizable Pancreas Segmentation via a Dual Self-Supervised Learning Framework
• Convolution-Weight-Distribution Assumption: Rethinking the Criteria of Channel Pruning
• End-to-End Trainable Multi-Instance Pose Estimation with Transformers
• Robustness of Accuracy Metric and its Inspirations in Learning with Noisy Labels
• SSA: Semantic Structure Aware Inference for Weakly Pixel-Wise Dense Predictions without Cost
• EcoNAS: Finding Proxies for Economical Neural Architecture Search
• Augmentation Strategies for Learning with Noisy Labels
• From Two to One: A New Scene Text Recognizer with Visual Language Modeling Network
• Semi-Supervised Domain Generalizable Person Re-Identification
• Adversarially Adaptive Normalization for Single Domain Generalization
• DartsReNet: Exploring new RNN cells in ReNet architectures
• Transportation mode recognition based on low-rate acceleration and location signals with an attention-based multiple-instance learning network
• Local and Global Context-and-Object-part-Aware Superpixel-based Data Augmentation for Deep Visual Recognition
• The Impact of Background Removal on Performance of Neural Networks for Fashion Image Classification and Segmentation
• Delta-STN: Efficient Bilevel Optimization for Neural Networks using Structured Response Jacobians
• SIRe-Networks: Convolutional Neural Networks Architectural Extension for Information Preservation via Skip/Residual Connections and Interlaced Auto-Encoders
• It Takes Two to Tango: Mixup for Deep Metric Learning
• Context Decoupling Augmentation for Weakly Supervised Semantic Segmentation
• Leave Zero Out: Towards a No-Cross-Validation Approach for Model Selection
• Google Landmark Recognition 2020 Competition Third Place Solution
• Deep Learning in Diabetic Foot Ulcers Detection: A Comprehensive Evaluation
• Few-shot Neural Architecture Search
• FeatMatch: Feature-Based Augmentation for Semi-Supervised Learning
• Balanced One-shot Neural Architecture Optimization
• Understanding Architectures Learnt by Cell-based Neural Architecture Search
• Online Hyper-parameter Learning for Auto-Augmentation Strategy
• Improving Neural Architecture Search Image Classifiers via Ensemble Learning
• Efficient Forward Architecture Search
• Learning Domain Invariant Representations by Joint Wasserstein Distance Minimization
• Hierarchical Weight Averaging for Deep Neural Networks
• FDFtNet: Facing Off Fake Images using Fake Detection Fine-tuning Network
• Learning Vision Transformer with Squeeze and Excitation for Facial Expression Recognition
• Prune and Replace NAS
• NASI: Label- and Data-agnostic Neural Architecture Search at Initialization
• Evaluating COPY-BLEND Augmentation for Low Level Vision Tasks
• Negative Data Augmentation
• Learning to Impute: A General Framework for Semi-supervised Learning
• Bit Error Robustness for Energy-Efficient DNN Accelerators
• Open Domain Generalization with Domain-Augmented Meta-Learning
• A Study on the Impact of Data Augmentation for Training Convolutional Neural Networks in the Presence of Noisy Labels
• Scale-aware Automatic Augmentation for Object Detection
• Adaptive Label Smoothing
• Low-cost and high-performance data augmentation for deep-learning-based skin lesion classification
• IRLAS: Inverse Reinforcement Learning for Architecture Search
• Training GANs with Stronger Augmentations via Contrastive Discriminator
• A Dual-Task Synergy-Driven Generalization Framework for Pancreatic Cancer Segmentation in CT Scans
• Circumventing Outliers of AutoAugment with Knowledge Distillation
• FakeMix Augmentation Improves Transparent Object Detection
• SnapMix: Semantically Proportional Mixing for Augmenting Fine-grained Data
• Automated Learning Rate Scheduler for Large-batch Training
• Measuring Visual Generalization in Continuous Control from Pixels
• A Bayesian Perspective of Convolutional Neural Networks through a Deconvolutional Generative Model
• A Three-Stage Self-Training Framework for Semi-Supervised Semantic Segmentation
• Sinkhorn Label Allocation: Semi-Supervised Classification via Annealed Self-Training
• Robustness and Transferability of Universal Attacks on Compressed Models
• A Simple Baseline for Semi-supervised Semantic Segmentation with Strong Data Augmentation
• A Realistic Evaluation of Semi-Supervised Learning for Fine-Grained Classification
• A Game-Theoretic Taxonomy of Visual Concepts in DNNs
• What Can be Seen is What You Get: Structure Aware Point Cloud Augmentation
• Rethinking Recurrent Neural Networks and Other Improvements for Image Classification
• ImmuNeCS: Neural Committee Search by an Artificial Immune System
• Unsupervised Pre-Training for 3D Leaf Instance Segmentation
• On the Reduction of Variance and Overestimation of Deep Q-Learning
• RobFR: Benchmarking Adversarial Robustness on Face Recognition
• Adversarial Learning of General Transformations for Data Augmentation
• When Human Pose Estimation Meets Robustness: Adversarial Algorithms and Benchmarks
• AugNet: End-to-End Unsupervised Visual Representation Learning with Image Augmentation
• Regularizing Neural Networks via Adversarial Model Perturbation
• CNNs Avoid Curse of Dimensionality by Learning on Patches
• BulletTrain: Accelerating Robust Neural Network Training via Boundary Example Mining
• Meta Approach to Data Augmentation Optimization
• Parallel Grid Pooling for Data Augmentation
• Efficient Decoupled Neural Architecture Search by Structure and Operation Sampling
• Trade-offs in Top-k Classification Accuracies on Losses for Deep Learning
• Inductive Transfer for Neural Architecture Optimization
• HM-NAS: Efficient Neural Architecture Search via Hierarchical Masking
• Rethinking the Number of Channels for the Convolutional Neural Network
• PAFNet: An Efficient Anchor-Free Object Detector Guidance
• Robust Semantic Segmentation with Superpixel-Mix
• Elsa: Energy-based learning for semi-supervised anomaly detection
• LRWR: Large-Scale Benchmark for Lip Reading in Russian language
• Focus Your Distribution: Coarse-to-Fine Non-Contrastive Learning for Anomaly Detection and Localization
• ReSmooth: Detecting and Utilizing OOD Samples when Training with Data Augmentation
• Deep Learning with Label Differential Privacy
• HAGAN: Hybrid Augmented Generative Adversarial Network for Medical Image Synthesis
• Series Saliency: Temporal Interpretation for Multivariate Time Series Forecasting
• Multi-objective Neural Architecture Search via Non-stationary Policy Gradient
• Improving Noise Robustness of an End-to-End Neural Model for Automatic Speech Recognition
• InstaNAS: Instance-aware Neural Architecture Search
• Augmentation Pathways Network for Visual Recognition
• MMF: Multi-Task Multi-Structure Fusion for Hierarchical Image Classification
• Catch-Up Mix: Catch-Up Class for Struggling Filters in CNN
• HourNAS: Extremely Fast Neural Architecture Search Through an Hourglass Lens
• Vision-Based Uncertainty-Aware Motion Planning based on Probabilistic Semantic Segmentation
• Can We Read Speech Beyond the Lips? Rethinking RoI Selection for Deep Visual Speech Recognition
• Almost Sure Convergence of Dropout Algorithms for Neural Networks
• Robust Object Detection under Occlusion with Context-Aware CompositionalNets
• Contrastive Neural Architecture Search with Neural Architecture Comparators
• You Look Twice: GaterNet for Dynamic Filter Selection in CNNs
• MarbleNet: Deep 1D Time-Channel Separable Convolutional Neural Network for Voice Activity Detection
• Augmentation Matters: A Mix-Paste Method for X-Ray Prohibited Item Detection under Noisy Annotations
• Integrating Large Circular Kernels into CNNs through Neural Architecture Search
• WSSOD: A New Pipeline for Weakly- and Semi-Supervised Object Detection
• Steerable Partial Differential Operators for Equivariant Neural Networks
• Convolutional Neural Networks with Dynamic Regularization
• Drop-Activation: Implicit Parameter Reduction and Harmonic Regularization
• A Preliminary Study on Data Augmentation of Deep Learning for Image Classification
• Learning to Rank Learning Curves
• Re-labeling ImageNet: from Single to Multi-Labels, from Global to Localized Labels
• Probabilistic Embeddings for Cross-Modal Retrieval
• Toward Learning Human-aligned Cross-domain Robust Models by Countering Misaligned Features
• An Asymptotically Optimal Multi-Armed Bandit Algorithm and Hyperparameter Optimization
• Repetitive Reprediction Deep Decipher for Semi-Supervised Learning
• Partial local entropy and anisotropy in deep weight spaces
• Geometry-Aware Gradient Algorithms for Neural Architecture Search
• ObjectAug: Object-level Data Augmentation for Semantic Image Segmentation
• Anti-aliasing Deep Image Classifiers using Novel Depth Adaptive Blurring and Activation Function
• EventDrop: data augmentation for event-based learning
• Deep Active Learning with Augmentation-based Consistency Estimation
• Multi-Objective Neural Architecture Search Based on Diverse Structures and Adaptive Recommendation
• An Empirical Study and Analysis on Open-Set Semi-Supervised Learning
• Diversity-Achieving Slow-DropBlock Network for Person Re-Identification
• Feature Binding with Category-Dependant MixUp for Semantic Segmentation and Adversarial Robustness
• Center-wise Local Image Mixture For Contrastive Representation Learning
• Bag of Tricks for Neural Architecture Search
• MutualNet: Adaptive ConvNet via Mutual Learning from Network Width and Resolution
• Does Data Augmentation Benefit from Split BatchNorms
• Analyzing Effects of Mixed Sample Data Augmentation on Model Interpretability
• ConDA: Unsupervised Domain Adaptation for LiDAR Segmentation via Regularized Domain Concatenation
• OnlineAugment: Online Data Augmentation with Less Domain Knowledge
• Disentanglement-based Cross-Domain Feature Augmentation for Effective Unsupervised Domain Adaptive Person Re-identification
• RankingMatch: Delving into Semi-Supervised Learning with Consistency Regularization and Ranking Loss
• Learning Optimal Data Augmentation Policies via Bayesian Optimization for Image Classification Tasks
• A New Look at Ghost Normalization
• 3rd Place Solution to Google Landmark Recognition Competition 2021
• Binarized Neural Architecture Search for Efficient Object Recognition
• Learning with Hierarchical Complement Objective
• Boosting Active Learning for Speech Recognition with Noisy Pseudo-labeled Samples
• Rethinking Performance Estimation in Neural Architecture Search
• Intra-Ensemble in Neural Networks
• Improving Robustness using Generated Data
• Attribute Mix: Semantic Data Augmentation for Fine Grained Recognition
• Lifelong Learning with Sketched Structural Regularization
• A Simple Dynamic Learning Rate Tuning Algorithm For Automated Training of DNNs
• CEKD:Cross Ensemble Knowledge Distillation for Augmented Fine-grained Data
• Pyramid Adversarial Training Improves ViT Performance
• Generative Interventions for Causal Learning
• Lesion-based Contrastive Learning for Diabetic Retinopathy Grading from Fundus Images
• How Not to Give a FLOP: Combining Regularization and Pruning for Efficient Inference
• ImageNet-21K Pretraining for the Masses
• ECACL: A Holistic Framework for Semi-Supervised Domain Adaptation
• Repaint: Improving the Generalization of Down-Stream Visual Tasks by Generating Multiple Instances of Training Examples
• Consistency Regularization for Deep Face Anti-Spoofing
• Understanding the Generalization Benefit of Model Invariance from a Data Perspective
• Histogram-Equalized Quantization for logic-gated Residual Neural Networks
• Causality-inspired Single-source Domain Generalization for Medical Image Segmentation
• Exploiting Operation Importance for Differentiable Neural Architecture Search
• Learning Temporally Invariant and Localizable Features via Data Augmentation for Video Recognition
• ConvNets for Counting: Object Detection of Transient Phenomena in Steelpan Drums
• Disturbance-immune Weight Sharing for Neural Architecture Search
• Uncertainty-guided Model Generalization to Unseen Domains
• Sub-word Level Lip Reading With Visual Attention
• Simple Modifications to Improve Tabular Neural Networks
• PriorityCut: Occlusion-guided Regularization for Warp-based Image Animation
• Separable Layers Enable Structured Efficient Linear Substitutions
• Unsupervised Document Embedding via Contrastive Augmentation
• Exploring 2D Data Augmentation for 3D Monocular Object Detection
• ApproxDARTS: Differentiable Neural Architecture Search with Approximate Multipliers
• Sequence-Level Mixed Sample Data Augmentation
• Robustness Out of the Box: Compositional Representations Naturally Defend Against Black-Box Patch Attacks
• WSAM: Visual Explanations from Style Augmentation as Adversarial Attacker and Their Influence in Image Classification
• HR-NAS: Searching Efficient High-Resolution Neural Architectures with Lightweight Transformers
• Augmented Cyclic Consistency Regularization for Unpaired Image-to-Image Translation
• Optimizing Neural Architecture Search using Limited GPU Time in a Dynamic Search Space: A Gene Expression Programming Approach
• Partial Graph Reasoning for Neural Network Regularization
• Binarized Neural Architecture Search
• Sim-To-Real Transfer for Miniature Autonomous Car Racing
• StackMix: A complementary Mix algorithm
• Evolving Neural Selection with Adaptive Regularization
• The Penalty Imposed by Ablated Data Augmentation
• PhytNet -- Tailored Convolutional Neural Networks for Custom Botanical Data
• Content-Aware Preserving Image Generation
• Defending Against Image Corruptions Through Adversarial Augmentations
• Amodal Segmentation through Out-of-Task and Out-of-Distribution Generalization with a Bayesian Model
• StyleAugment: Learning Texture De-biased Representations by Style Augmentation without Pre-defined Textures
• How Does Supernet Help in Neural Architecture Search?
• Neighborhood-Aware Neural Architecture Search
• Random Shadows and Highlights: A new data augmentation method for extreme lighting conditions
• Data augmentation and image understanding
• Improving Few-Shot Learning with Auxiliary Self-Supervised Pretext Tasks
• Adaptive Low-Rank Factorization to regularize shallow and deep neural networks
• Improved Image Augmentation for Convolutional Neural Networks by Copyout and CopyPairing
• Unsupervised Deep Metric Learning with Transformed Attention Consistency and Contrastive Clustering Loss
• G-DARTS-A: Groups of Channel Parallel Sampling with Attention
• Class-Wise Difficulty-Balanced Loss for Solving Class-Imbalance
• Dynamic Hierarchical Mimicking Towards Consistent Optimization Objectives
• Consistency Regularization Can Improve Robustness to Label Noise
• Anti-Bandit Neural Architecture Search for Model Defense
• SSMD: Semi-Supervised Medical Image Detection with Adaptive Consistency and Heterogeneous Perturbation
• Genetic Network Architecture Search
• Direct Differentiable Augmentation Search
• Analysis of skin lesion images with deep learning
• Cooperative Training and Latent Space Data Augmentation for Robust Medical Image Segmentation
• DropFilter: A Novel Regularization Method for Learning Convolutional Neural Networks
• Look at here : Utilizing supervision to attend subtle key regions
• Black-Box Dissector: Towards Erasing-based Hard-Label Model Stealing Attack
• Regularized Evolutionary Population-Based Training
• ResiliNet: Failure-Resilient Inference in Distributed Neural Networks
• Reweighting Augmented Samples by Minimizing the Maximal Expected Loss
• Efficient Method for Categorize Animals in the Wild
• Dense xUnit Networks
• Neural Audio Fingerprint for High-specific Audio Retrieval based on Contrastive Learning
• A Unified Game-Theoretic Interpretation of Adversarial Robustness
• Towards Efficient Convolutional Network Models with Filter Distribution Templates
• Data augmentation with Mobius transformations
• Beyond Dropout: Feature Map Distortion to Regularize Deep Neural Networks
• Improving Augmentation and Evaluation Schemes for Semantic Image Synthesis
• Neural Function Modules with Sparse Arguments: A Dynamic Approach to Integrating Information across Layers
• Hypernetwork-Based Augmentation
• Semi-supervised Quality Evaluation of Colonoscopy Procedures
• Improving Polyphonic Sound Event Detection on Multichannel Recordings with the Sørensen-Dice Coefficient Loss and Transfer Learning
• Deep Feature Response Discriminative Calibration
• Fine-tuning Handwriting Recognition systems with Temporal Dropout
• Margin-Based Regularization and Selective Sampling in Deep Neural Networks
• On the Orthogonality of Knowledge Distillation with Other Techniques: From an Ensemble Perspective
• Humble Teachers Teach Better Students for Semi-Supervised Object Detection
• Multi-Plateau Ensemble for Endoscopic Artefact Segmentation and Detection
• Hippo: Taming Hyper-parameter Optimization of Deep Learning with Stage Trees
• A Unified Mixture-View Framework for Unsupervised Representation Learning
• KoSpeech: Open-Source Toolkit for End-to-End Korean Speech Recognition
• Selective sampling for accelerating training of deep neural networks
• Augmentation Inside the Network
• RaftMLP: How Much Can Be Done Without Attention and with Less Spatial Locality?
• MS-DARTS: Mean-Shift Based Differentiable Architecture Search
• ParamCrop: Parametric Cubic Cropping for Video Contrastive Learning
• Instance based Generalization in Reinforcement Learning
• Multi-view Feature Augmentation with Adaptive Class Activation Mapping
• Multi-objective Neural Architecture Search with Almost No Training
• Unsupervised Visual Representation Learning by Tracking Patches in Video
• EDDA: Explanation-driven Data Augmentation to Improve Explanation Faithfulness
• Self-paced Data Augmentation for Training Neural Networks
• Boosting the Performance of Semi-Supervised Learning with Unsupervised Clustering
• Robust Instance Segmentation through Reasoning about Multi-Object Occlusion
• Local Patch AutoAugment with Multi-Agent Collaboration
• Adversarial AutoMixup
• ReRankMatch: Semi-Supervised Learning with Semantics-Oriented Similarity Representation
• Knowledge Transfer Based Fine-grained Visual Classification
• Out-distribution aware Self-training in an Open World Setting
• MetaAugment: Sample-Aware Data Augmentation Policy Learning
• Effective Regularization Through Loss-Function Metalearning
• How Data Augmentation affects Optimization for Linear Regression
• Mixup Without Hesitation
• Medical Image Segmentation with Limited Supervision: A Review of Deep Network Models
• Efficient Model Performance Estimation via Feature Histories
• False Negative Distillation and Contrastive Learning for Personalized Outfit Recommendation
• SMCL: Saliency Masked Contrastive Learning for Long-tailed Recognition
• Improving compute efficacy frontiers with SliceOut
• Progressive Multi-stage Feature Mix for Person Re-Identification
• Neural Parameter Allocation Search
• Unsupervised Person Re-identification via Simultaneous Clustering and Consistency Learning
• Explainability-aided Domain Generalization for Image Classification
• Data Augmentation via Mixed Class Interpolation using Cycle-Consistent Generative Adversarial Networks Applied to Cross-Domain Imagery
• Gradient-based Data Augmentation for Semi-Supervised Learning
• LA3: Efficient Label-Aware AutoAugment
• Temporally Resolution Decrement: Utilizing the Shape Consistency for Higher Computational Efficiency
• Empirical Perspectives on One-Shot Semi-supervised Learning
• Do CNNs Encode Data Augmentations?
• Dual-attention Guided Dropblock Module for Weakly Supervised Object Localization
• Compositional Convolutional Neural Networks: A Deep Architecture with Innate Robustness to Partial Occlusion
• Upgrading the Newsroom: An Automated Image Selection System for News Articles
• SuperpixelGridCut, SuperpixelGridMean and SuperpixelGridMix Data Augmentation
• Self-supervised Lesion Change Detection and Localisation in Longitudinal Multiple Sclerosis Brain Imaging
• Channel DropBlock: An Improved Regularization Method for Fine-Grained Visual Classification
• Adversarial Policy Gradient for Deep Learning Image Augmentation
• On the Accuracy of CRNNs for Line-Based OCR: A Multi-Parameter Evaluation
• Pointwise visual field estimation from optical coherence tomography in glaucoma: a structure-function analysis using deep learning
• Scale Calibrated Training: Improving Generalization of Deep Networks via Scale-Specific Normalization
• Improved Robustness to Open Set Inputs via Tempered Mixup
• Dynamic Routing Networks
• EENA: Efficient Evolution of Neural Architecture
• Searching for A Robust Neural Architecture in Four GPU Hours
• Parametric Contrastive Learning
• Understanding the Disharmony between Weight Normalization Family and Weight Decay: $\epsilon -$shifted $L_2$ Regularizer
• A Visual Domain Transfer Learning Approach for Heartbeat Sound Classification
• Seismic horizon detection with neural networks
• ADWPNAS: Architecture-Driven Weight Prediction for Neural Architecture Search
• Deformation Robust Roto-Scale-Translation Equivariant CNNs
• Cascade Attentive Dropout for Weakly Supervised Object Detection
• AlphaMatch: Improving Consistency for Semi-supervised Learning with Alpha-divergence
• Unsupervised Object Learning via Common Fate
• DeepRepair: Style-Guided Repairing for DNNs in the Real-world Operational Environment
• ROI Regularization for Semi-supervised and Supervised Learning
• The NTNU System at the Interspeech 2020 Non-Native Children's Speech ASR Challenge
• Stochastic Region Pooling: Make Attention More Expressive
• CP-NAS: Child-Parent Neural Architecture Search for Binary Neural Networks
• Information based Deep Clustering: An experimental study
• ShuffleBlock: Shuffle to Regularize Deep Convolutional Neural Networks
• Farkas layers: don't shift the data, fix the geometry
• Harnessing Perceptual Adversarial Patches for Crowd Counting
• Itsy Bitsy SpiderNet: Fully Connected Residual Network for Fraud Detection
• Robust Model-based Face Reconstruction through Weakly-Supervised Outlier Segmentation
• Spatially Attentive Output Layer for Image Classification
• Introduce the Result Into Self-Attention
• Free Lunch for Co-Saliency Detection: Context Adjustment
• Induced Feature Selection by Structured Pruning
• Optimizing Data Augmentation Policy Through Random Unidimensional Search
• SwGridNet: A Deep Convolutional Neural Network based on Grid Topology for Image Classification
• Coarse-To-Fine And Cross-Lingual ASR Transfer
• RotationOut as a Regularization Method for Neural Network
• The Robust Semantic Segmentation UNCV2023 Challenge Results
• Inspect Transfer Learning Architecture with Dilated Convolution
• Unified Multifaceted Feature Learning for Person Re-Identification
• ReMix: Towards Image-to-Image Translation with Limited Data
• Role of Mixup in Topological Persistence Based Knowledge Distillation for Wearable Sensor Data
• Unsupervised Learning of Depth and Depth-of-Field Effect from Natural Images with Aperture Rendering Generative Adversarial Networks
• Parametric Variational Linear Units (PVLUs) in Deep Convolutional Networks
• Adversarial Examples as an Input-Fault Tolerance Problem
• Adaptive Hierarchical Hyper-gradient Descent
• A novel multiple instance learning framework for COVID-19 severity assessment via data augmentation and self-supervised learning
• BNAS-v2: Memory-efficient and Performance-collapse-prevented Broad Neural Architecture Search
• MSR-DARTS: Minimum Stable Rank of Differentiable Architecture Search
• In-Loop Meta-Learning with Gradient-Alignment Reward
• Split-and-Bridge: Adaptable Class Incremental Learning within a Single Neural Network
• Global and Local Texture Randomization for Synthetic-to-Real Semantic Segmentation
• Intra-class Patch Swap for Self-Distillation
• Neural Architecture Search via Combinatorial Multi-Armed Bandit
• ConAM: Confidence Attention Module for Convolutional Neural Networks
• Implicit Label Augmentation on Partially Annotated Clips via Temporally-Adaptive Features Learning
• What Can Style Transfer and Paintings Do For Model Robustness?
• Towards More Efficient and Effective Inference: The Joint Decision of Multi-Participants
• Evolving Neural Architecture Using One Shot Model
• Point Cloud Segmentation for Autonomous Clip Positioning in Laparoscopic Cholecystectomy on a Phantom
• MobileFace: 3D Face Reconstruction with Efficient CNN Regression
• Towards Data-Free Domain Generalization
• Recent Advances in Large Margin Learning
• Fine-Grained AutoAugmentation for Multi-Label Classification
• Improve SGD Training via Aligning Mini-batches
• GPNAS: A Neural Network Architecture Search Framework Based on Graphical Predictor
• FROST: Faster and more Robust One-shot Semi-supervised Training
• Image classification and retrieval with random depthwise signed convolutional neural networks
• Missingness Augmentation: A General Approach for Improving Generative Imputation Models
• Bootstrap Your Object Detector via Mixed Training
• Selective Output Smoothing Regularization: Regularize Neural Networks by Softening Output Distributions
• SelectScale: Mining More Patterns from Images via Selective and Soft Dropout
• Pathology-Aware Generative Adversarial Networks for Medical Image Augmentation
• Credal Self-Supervised Learning
• Self-supervision of Feature Transformation for Further Improving Supervised Learning
• 3-Dimensional Deep Learning with Spatial Erasing for Unsupervised Anomaly Segmentation in Brain MRI
• Safe Augmentation: Learning Task-Specific Transformations from Data
• Predicting Driver Intention Using Deep Neural Network
• Intra-Model Collaborative Learning of Neural Networks
• Unsupervised Representation Learning with Laplacian Pyramid Auto-encoders
• Stochastic Batch Augmentation with An Effective Distilled Dynamic Soft Label Regularizer
• Robust Face Verification via Disentangled Representations
• Sequential Feature Filtering Classifier
• Deep Goal-Oriented Clustering
• Indoor Scene Recognition in 3D
• Go Small and Similar: A Simple Output Decay Brings Better Performance
• AI Enlightens Wireless Communication: Analyses, Solutions and Opportunities on CSI Feedback