Densely Connected Convolutional Networks

Abstract

Recent work has shown that convolutional networks can be substantially deeper, more accurate, and efficient to train if they contain shorter connections between layers close to the input and those close to the output. In this paper, we embrace this observation and introduce the Dense Convolutional Network (DenseNet), which connects each layer to every other layer in a feed-forward fashion. Whereas traditional convolutional networks with L layers have L connections - one between each layer and its subsequent layer - our network has L(L+1)/2 direct connections. For each layer, the feature-maps of all preceding layers are used as inputs, and its own feature-maps are used as inputs into all subsequent layers. DenseNets have several compelling advantages: they alleviate the vanishing-gradient problem, strengthen feature propagation, encourage feature reuse, and substantially reduce the number of parameters. We evaluate our proposed architecture on four highly competitive object recognition benchmark tasks (CIFAR-10, CIFAR-100, SVHN, and ImageNet). DenseNets obtain significant improvements over the state-of-the-art on most of them, whilst requiring less computation to achieve high performance. Code and pre-trained models are available at https://github.com/liuzhuang13/DenseNet .

CVPR 2017

References in corpus (12)

• Striving for Simplicity: The All Convolutional Net
• FitNets: Hints for Thin Deep Nets
• Wide Residual Networks
• Going Deeper with Convolutions
• Fully Convolutional Networks for Semantic Segmentation
• Resnet in Resnet: Generalizing Residual Architectures
• FractalNet: Ultra-Deep Neural Networks without Residuals
• Bridging the Gaps Between Residual Learning, Recurrent Neural Networks and Visual Cortex
• AdaNet: Adaptive Structural Learning of Artificial Neural Networks
• Memory-Efficient Implementation of DenseNets
• Deeply-Fused Nets
• Pedestrian Detection with Unsupervised Multi-Stage Feature Learning

Cited by in corpus (440)

• Neural Architecture Search with Reinforcement Learning
• ResUNet-a: a deep learning framework for semantic segmentation of remotely sensed data
• SGDR: Stochastic Gradient Descent with Warm Restarts
• CheXNet: Radiologist-Level Pneumonia Detection on Chest X-Rays with Deep Learning
• fastai: A Layered API for Deep Learning
• COVID-19 Image Data Collection: Prospective Predictions Are the Future
• Data Augmentation Generative Adversarial Networks
• Bayesian Deep Convolutional Encoder-Decoder Networks for Surrogate Modeling and Uncertainty Quantification
• BEGAN: Boundary Equilibrium Generative Adversarial Networks
• Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks
• FractalNet: Ultra-Deep Neural Networks without Residuals
• Efficient Neural Architecture Search via Parameter Sharing
• Automatic Ship Detection of Remote Sensing Images from Google Earth in Complex Scenes Based on Multi-Scale Rotation Dense Feature Pyramid Networks
• Deep Clustering for Unsupervised Learning of Visual Features
• Recurrent Residual Convolutional Neural Network based on U-Net (R2U-Net) for Medical Image Segmentation
• The History Began from AlexNet: A Comprehensive Survey on Deep Learning Approaches
• Advancements in Image Classification using Convolutional Neural Network
• Quasi-Recurrent Neural Networks
• Shake-Shake regularization
• Machine Learning for Wireless Communications in the Internet of Things: A Comprehensive Survey
• LocalViT: Analyzing Locality in Vision Transformers
• Deep Learning Enables Automatic Detection and Segmentation of Brain Metastases on Multi-Sequence MRI
• Neural Photo Editing with Introspective Adversarial Networks
• Diagnose like a Radiologist: Attention Guided Convolutional Neural Network for Thorax Disease Classification
• Exploring the Regularity of Sparse Structure in Convolutional Neural Networks
• Hierarchical Representations for Efficient Architecture Search
• Spectral Norm Regularization for Improving the Generalizability of Deep Learning
• A Closer Look at Spatiotemporal Convolutions for Action Recognition
• Early-detection and classification of live bacteria using time-lapse coherent imaging and deep learning
• Learning to diagnose from scratch by exploiting dependencies among labels
• Mitigating Adversarial Effects Through Randomization
• Regularization for Deep Learning: A Taxonomy
• Simple And Efficient Architecture Search for Convolutional Neural Networks
• Threat of Adversarial Attacks on Deep Learning in Computer Vision: A Survey
• Progressive Neural Architecture Search
• Scaling Limits of Wide Neural Networks with Weight Sharing: Gaussian Process Behavior, Gradient Independence, and Neural Tangent Kernel Derivation
• Beyond Finite Layer Neural Networks: Bridging Deep Architectures and Numerical Differential Equations
• Path-Level Network Transformation for Efficient Architecture Search
• Beyond Data and Model Parallelism for Deep Neural Networks
• Bag of Tricks for Image Classification with Convolutional Neural Networks
• Context Encoding for Semantic Segmentation
• Margin Sample Mining Loss: A Deep Learning Based Method for Person Re-identification
• Galaxy Morphology Classification with Deep Convolutional Neural Networks
• ECG Arrhythmia Classification Using Transfer Learning from 2-Dimensional Deep CNN Features
• Deep Chain HDRI: Reconstructing a High Dynamic Range Image from a Single Low Dynamic Range Image
• Adversarial Examples, Uncertainty, and Transfer Testing Robustness in Gaussian Process Hybrid Deep Networks
• XNOR-Net++: Improved Binary Neural Networks
• Gated-SCNN: Gated Shape CNNs for Semantic Segmentation
• Super-Resolution via Deep Learning
• AdaNet: Adaptive Structural Learning of Artificial Neural Networks
• What Do We Understand About Convolutional Networks?
• To understand deep learning we need to understand kernel learning
• Efficient Two-Stream Network for Violence Detection Using Separable Convolutional LSTM
• Density-aware Single Image De-raining using a Multi-stream Dense Network
• Evaluating Prediction-Time Batch Normalization for Robustness under Covariate Shift
• Re-ID done right: towards good practices for person re-identification
• The Power of Sparsity in Convolutional Neural Networks
• The Pitfalls of Simplicity Bias in Neural Networks
• Measuring the Algorithmic Efficiency of Neural Networks
• Exploiting deep residual networks for human action recognition from skeletal data
• Predicting Lung Nodule Malignancies by Combining Deep Convolutional Neural Network and Handcrafted Features
• Peephole: Predicting Network Performance Before Training
• Self-Driving Car Steering Angle Prediction Based on Image Recognition
• Interleaved Group Convolutions for Deep Neural Networks
• A Deep Learning-based Reconstruction of Cosmic Ray-induced Air Showers
• Learning Spatio-Temporal Features with 3D Residual Networks for Action Recognition
• APE-GAN: Adversarial Perturbation Elimination with GAN
• New high-quality strong lens candidates with deep learning in the Kilo Degree Survey
• DiracNets: Training Very Deep Neural Networks Without Skip-Connections
• Neural Networks and Quantum Field Theory
• A Counter-Forensic Method for CNN-Based Camera Model Identification
• UnsuperPoint: End-to-end Unsupervised Interest Point Detector and Descriptor
• Do Convolutional Networks need to be Deep for Text Classification ?
• NSML: A Machine Learning Platform That Enables You to Focus on Your Models
• Massive Exploration of Neural Machine Translation Architectures
• Multi-Modal Emotion recognition on IEMOCAP Dataset using Deep Learning
• The One Hundred Layers Tiramisu: Fully Convolutional DenseNets for Semantic Segmentation
• Composition Loss for Counting, Density Map Estimation and Localization in Dense Crowds
• A Genetic Programming Approach to Designing Convolutional Neural Network Architectures
• Acoustic Anomaly Detection for Machine Sounds based on Image Transfer Learning
• Data augmentation using learned transformations for one-shot medical image segmentation
• Reading Scene Text with Attention Convolutional Sequence Modeling
• Lets keep it simple, Using simple architectures to outperform deeper and more complex architectures
• Detecting hip fractures with radiologist-level performance using deep neural networks
• DeepStreaks: identifying fast-moving objects in the Zwicky Transient Facility data with deep learning
• Automatic 3D Cardiovascular MR Segmentation with Densely-Connected Volumetric ConvNets
• A Survey on Deep Learning Methods for Robot Vision
• Adversarial Examples in Modern Machine Learning: A Review
• Two-Layer Neural Networks for Partial Differential Equations: Optimization and Generalization Theory
• Deep Convolutional Neural Network Design Patterns
• WasteNet: Waste Classification at the Edge for Smart Bins
• Activation Maximization Generative Adversarial Nets
• Rapid Identification of Strongly Lensed Gravitational-Wave Events with Machine Learning
• Deep Transfer Learning Methods for Colon Cancer Classification in Confocal Laser Microscopy Images
• Lesion Segmentation in Ultrasound Using Semi-pixel-wise Cycle Generative Adversarial Nets
• Jasper: An End-to-End Convolutional Neural Acoustic Model
• Medical Imaging and Computational Image Analysis in COVID-19 Diagnosis: A Review
• LSTM Based Music Generation System
• Tensor Programs I: Wide Feedforward or Recurrent Neural Networks of Any Architecture are Gaussian Processes
• Direct Detection of Dark Matter Substructure in Strong Lens Images with Convolutional Neural Networks
• Optimization on Submanifolds of Convolution Kernels in CNNs
• Understanding the Disharmony between Dropout and Batch Normalization by Variance Shift
• Multi-Residual Networks: Improving the Speed and Accuracy of Residual Networks
• Densely Connected CNNs for Bird Audio Detection
• Relation-Aware Global Attention for Person Re-identification
• Deep Interest Network for Click-Through Rate Prediction
• Patch-based Progressive 3D Point Set Upsampling
• PolyNet: A Pursuit of Structural Diversity in Very Deep Networks
• U-Net: Machine Reading Comprehension with Unanswerable Questions
• Adversarially Robust Neural Architectures
• DCFNet: Deep Neural Network with Decomposed Convolutional Filters
• EXTD: Extremely Tiny Face Detector via Iterative Filter Reuse
• Stacked Conditional Generative Adversarial Networks for Jointly Learning Shadow Detection and Shadow Removal
• General audio tagging with ensembling convolutional neural network and statistical features
• Towards High Performance Video Object Detection for Mobiles
• Deep Gamblers: Learning to Abstain with Portfolio Theory
• MultiGrain: a unified image embedding for classes and instances
• Temporal Generative Adversarial Nets with Singular Value Clipping
• Deep Verifier Networks: Verification of Deep Discriminative Models with Deep Generative Models
• Deep Convolutional Neural Networks with Merge-and-Run Mappings
• Adversarial Examples Make Strong Poisons
• Sparse Attentive Backtracking: Temporal CreditAssignment Through Reminding
• A Non-Technical Survey on Deep Convolutional Neural Network Architectures
• Maintaining Discrimination and Fairness in Class Incremental Learning
• Noisy Softmax: Improving the Generalization Ability of DCNN via Postponing the Early Softmax Saturation
• Deep Pyramidal Residual Networks
• Universal Adversarial Perturbations: A Survey
• Class-incremental Learning via Deep Model Consolidation
• Hyperparameter Optimization: A Spectral Approach
• COVID-19 Screening Using Residual Attention Network an Artificial Intelligence Approach
• IGCV3: Interleaved Low-Rank Group Convolutions for Efficient Deep Neural Networks
• Wide Compression: Tensor Ring Nets
• Quark-Gluon Jet Discrimination Using Convolutional Neural Networks
• Relational Learning between Multiple Pulmonary Nodules via Deep Set Attention Transformers
• Memory Attention Networks for Skeleton-based Action Recognition
• Deep Back-Projection Networks For Super-Resolution
• Deep Transfer Learning for Static Malware Classification
• Image to Image Translation for Domain Adaptation
• Defense against Universal Adversarial Perturbations
• Graph Convolution: A High-Order and Adaptive Approach
• Spatially Adaptive Computation Time for Residual Networks
• Rethinking Feature Distribution for Loss Functions in Image Classification
• Predify: Augmenting deep neural networks with brain-inspired predictive coding dynamics
• DenseNet Models for Tiny ImageNet Classification
• Imaging through glass diffusers using densely connected convolutional networks
• NeuPDE: Neural Network Based Ordinary and Partial Differential Equations for Modeling Time-Dependent Data
• DARTS: DenseUnet-based Automatic Rapid Tool for brain Segmentation
• Knowledge Concentration: Learning 100K Object Classifiers in a Single CNN
• Multimodal Emotion Recognition for One-Minute-Gradual Emotion Challenge
• Applications of machine learning in gravitational wave research with current interferometric detectors
• High-Quality Face Image SR Using Conditional Generative Adversarial Networks
• Deep Neural Network Architectures for Modulation Classification
• SIXray : A Large-scale Security Inspection X-ray Benchmark for Prohibited Item Discovery in Overlapping Images
• Image Super-Resolution via Dual-State Recurrent Networks
• Explaining Convolutional Neural Networks using Softmax Gradient Layer-wise Relevance Propagation
• Self-Supervised Variational Auto-Encoders
• Analysis and Design of Convolutional Networks via Hierarchical Tensor Decompositions
• Agile Amulet: Real-Time Salient Object Detection with Contextual Attention
• Fast and Accurate Sparse Coding of Visual Stimuli with a Simple, Ultra-Low-Energy Spiking Architecture
• Deep Pyramidal Residual Networks with Separated Stochastic Depth
• Self-Attention Capsule Networks for Object Classification
• Not All Ops Are Created Equal!
• 3D Tomographic Pattern Synthesis for Enhancing the Quantification of COVID-19
• Multi-Scale Attention with Dense Encoder for Handwritten Mathematical Expression Recognition
• Highrisk Prediction from Electronic Medical Records via Deep Attention Networks
• Revisiting Single Image Depth Estimation: Toward Higher Resolution Maps with Accurate Object Boundaries
• Parallel Separable 3D Convolution for Video and Volumetric Data Understanding
• Parle: parallelizing stochastic gradient descent
• Few-Shot Image Recognition by Predicting Parameters from Activations
• Improved YOLOv5s model for key components detection of power transmission lines
• Protein Secondary Structure Prediction Using Deep Multi-scale Convolutional Neural Networks and Next-Step Conditioning
• Multi-scale Multi-band DenseNets for Audio Source Separation
• Next-Step Conditioned Deep Convolutional Neural Networks Improve Protein Secondary Structure Prediction
• Finding Better Topologies for Deep Convolutional Neural Networks by Evolution
• CovidExpert: A Triplet Siamese Neural Network framework for the detection of COVID-19
• An interpretable classifier for high-resolution breast cancer screening images utilizing weakly supervised localization
• MetaSAug: Meta Semantic Augmentation for Long-Tailed Visual Recognition
• Advances in Deep Space Exploration via Simulators & Deep Learning
• Learning Compact Recurrent Neural Networks with Block-Term Tensor Decomposition
• Exploring Extended Reality with ILLIXR: A New Playground for Architecture Research
• Adjusting for Dropout Variance in Batch Normalization and Weight Initialization
• Training Deeper Neural Machine Translation Models with Transparent Attention
• CheXphoto: 10,000+ Photos and Transformations of Chest X-rays for Benchmarking Deep Learning Robustness
• A Statistical Approach to Assessing Neural Network Robustness
• Breaking Batch Normalization for better explainability of Deep Neural Networks through Layer-wise Relevance Propagation
• Single Image 3D Hand Reconstruction with Mesh Convolutions
• MatchboxNet: 1D Time-Channel Separable Convolutional Neural Network Architecture for Speech Commands Recognition
• Data Hiding with Deep Learning: A Survey Unifying Digital Watermarking and Steganography
• Degeneration in VAE: in the Light of Fisher Information Loss
• BitPruning: Learning Bitlengths for Aggressive and Accurate Quantization
• Neural Sign Language Translation based on Human Keypoint Estimation
• FD-MobileNet: Improved MobileNet with a Fast Downsampling Strategy
• Constructing Deep Neural Networks by Bayesian Network Structure Learning
• The Deep Ritz method: A deep learning-based numerical algorithm for solving variational problems
• Single-Shot Object Detection with Enriched Semantics
• Neural Transformers for Intraductal Papillary Mucosal Neoplasms (IPMN) Classification in MRI images
• Dense Dilated UNet: Deep Learning for 3D Photoacoustic Tomography Image Reconstruction
• Dual Attention Suppression Attack: Generate Adversarial Camouflage in Physical World
• Improving Video Generation for Multi-functional Applications
• Fine-Grained Neural Architecture Search
• UTRNet: High-Resolution Urdu Text Recognition In Printed Documents
• Robot-Assisted Feeding: Generalizing Skewering Strategies across Food Items on a Realistic Plate
• On the impact of selected modern deep-learning techniques to the performance and celerity of classification models in an experimental high-energy physics use case
• Artificial Intelligence Assisted Inversion (AIAI) of Synthetic Type Ia Supernova Spectra
• Deep Learning for the Digital Pathologic Diagnosis of Cholangiocarcinoma and Hepatocellular Carcinoma: Evaluating the Impact of a Web-based Diagnostic Assistant
• Non-locally Enhanced Encoder-Decoder Network for Single Image De-raining
• SESR: Single Image Super Resolution with Recursive Squeeze and Excitation Networks
• DenseRAN for Offline Handwritten Chinese Character Recognition
• Improved Stereo Matching with Constant Highway Networks and Reflective Confidence Learning
• Liver Lesion Segmentation with slice-wise 2D Tiramisu and Tversky loss function
• Learning to Write Stylized Chinese Characters by Reading a Handful of Examples
• Improving Deep Neural Network with Multiple Parametric Exponential Linear Units
• Detection of Einstein Telescope gravitational wave signals from binary black holes using deep learning
• Combining Image Features and Patient Metadata to Enhance Transfer Learning
• QUANOS- Adversarial Noise Sensitivity Driven Hybrid Quantization of Neural Networks
• Evaluating Generalization Ability of Convolutional Neural Networks and Capsule Networks for Image Classification via Top-2 Classification
• Inferring Point Clouds from Single Monocular Images by Depth Intermediation
• A Multi-Stage Multi-Task Neural Network for Aerial Scene Interpretation and Geolocalization
• Preventing Unauthorized Use of Proprietary Data: Poisoning for Secure Dataset Release
• A linearized framework and a new benchmark for model selection for fine-tuning
• Optimize transfer learning for lung diseases in bronchoscopy using a new concept: sequential fine-tuning
• Autonomous Deep Learning: A Genetic DCNN Designer for Image Classification
• Full deep neural network training on a pruned weight budget
• Neural Architecture Search using Deep Neural Networks and Monte Carlo Tree Search
• DOCTOR: A Simple Method for Detecting Misclassification Errors
• Reconstruction of 3-D Atomic Distortions from Electron Microscopy with Deep Learning
• Dual Path Networks for Multi-Person Human Pose Estimation
• Automatic Group Cohesiveness Detection With Multi-modal Features
• The End-of-End-to-End: A Video Understanding Pentathlon Challenge (2020)
• Grid-GCN for Fast and Scalable Point Cloud Learning
• Sparse Attentive Backtracking: Long-Range Credit Assignment in Recurrent Networks
• IrisNet: Deep Learning for Automatic and Real-time Tongue Contour Tracking in Ultrasound Video Data using Peripheral Vision
• Face Detection with Feature Pyramids and Landmarks
• An Interpretable Deep Hierarchical Semantic Convolutional Neural Network for Lung Nodule Malignancy Classification
• Knowledge-based Radiation Treatment Planning: A Data-driven Method Survey
• SYMOG: learning symmetric mixture of Gaussian modes for improved fixed-point quantization
• Generalized Negative Correlation Learning for Deep Ensembling
• Improving GAN Training via Binarized Representation Entropy (BRE) Regularization
• A Simple Cache Model for Image Recognition
• SSN: Learning Sparse Switchable Normalization via SparsestMax
• Re-identification = Retrieval + Verification: Back to Essence and Forward with a New Metric
• Benanza: Automatic $\mu$Benchmark Generation to Compute "Lower-bound" Latency and Inform Optimizations of Deep Learning Models on GPUs
• Automatic breast cancer grading in lymph nodes using a deep neural network
• Surface Defect Classification in Real-Time Using Convolutional Neural Networks
• Detection of Premature Ventricular Contractions Using Densely Connected Deep Convolutional Neural Network with Spatial Pyramid Pooling Layer
• AVA: Adversarial Vignetting Attack against Visual Recognition
• Natural Language Inference over Interaction Space: ICLR 2018 Reproducibility Report
• Better the Devil you Know: An Analysis of Evasion Attacks using Out-of-Distribution Adversarial Examples
• Fast Neural Architecture Construction using EnvelopeNets
• Residual Squeeze VGG16
• ResFPN: Residual Skip Connections in Multi-Resolution Feature Pyramid Networks for Accurate Dense Pixel Matching
• AID++: An Updated Version of AID on Scene Classification
• State Classification with CNN
• Multiscale Hierarchical Convolutional Networks
• ModuleNet: Knowledge-inherited Neural Architecture Search
• Analysis on DeepLabV3+ Performance for Automatic Steel Defects Detection
• Crowd Transformer Network
• A deep learning-based method for prostate segmentation in T2-weighted magnetic resonance imaging
• ESAI: Efficient Split Artificial Intelligence via Early Exiting Using Neural Architecture Search
• Do Vision Models Encode Object-Level Semantic Relatedness? A Cognitive Psychology-Inspired Benchmark
• Deep Learning for Moving Blockage Prediction using Real Millimeter Wave Measurements
• A Novel lightweight Convolutional Neural Network, ExquisiteNetV2
• Deep Learning Assessment of Tumor Proliferation in Breast Cancer Histological Images
• Unsupervised Sparse Dirichlet-Net for Hyperspectral Image Super-Resolution
• Understanding Deep Learning Generalization by Maximum Entropy
• Locally Adaptive Learning Loss for Semantic Image Segmentation
• Beyond Domain Adaptation: Unseen Domain Encapsulation via Universal Non-volume Preserving Models
• Data Efficient Stagewise Knowledge Distillation
• Visualizing the Hidden Features of Galaxy Morphology with Machine Learning
• Apparel-invariant Feature Learning for Apparel-changed Person Re-identification
• Noisy Computations during Inference: Harmful or Helpful?
• Multigrid Neural Architectures
• The Resistance to Label Noise in K-NN and DNN Depends on its Concentration
• Train on Validation: Squeezing the Data Lemon
• Report on UG^2+ Challenge Track 1: Assessing Algorithms to Improve Video Object Detection and Classification from Unconstrained Mobility Platforms
• MMGCN: Multimodal Fusion via Deep Graph Convolution Network for Emotion Recognition in Conversation
• AI Oriented Large-Scale Video Management for Smart City: Technologies, Standards and Beyond
• Beyond Forward Shortcuts: Fully Convolutional Master-Slave Networks (MSNets) with Backward Skip Connections for Semantic Segmentation
• DelugeNets: Deep Networks with Efficient and Flexible Cross-layer Information Inflows
• DeepLung: 3D Deep Convolutional Nets for Automated Pulmonary Nodule Detection and Classification
• Ultrasound-Guided Robotic Navigation with Deep Reinforcement Learning
• MutualNet: Adaptive ConvNet via Mutual Learning from Network Width and Resolution
• Generalized Bayesian Posterior Expectation Distillation for Deep Neural Networks
• Structural Residual Learning for Single Image Rain Removal
• End to End Recognition System for Recognizing Offline Unconstrained Vietnamese Handwriting
• Functional Gradient Boosting based on Residual Network Perception
• Stacked Neural Networks for end-to-end ciliary motion analysis
• Dual Encoder Fusion U-Net (DEFU-Net) for Cross-manufacturer Chest X-ray Segmentation
• Evolving Deep Neural Networks by Multi-objective Particle Swarm Optimization for Image Classification
• Deep Learning for Low-Dose CT Denoising
• Predicting Localized Primordial Star Formation with Deep Convolutional Neural Networks
• Deep joint rain and haze removal from single images
• Dense Information Flow for Neural Machine Translation
• Identifying Most Walkable Direction for Navigation in an Outdoor Environment
• Semi-Supervised Noisy Student Pre-training on EfficientNet Architectures for Plant Pathology Classification
• Deep Learning Seismic Substructure Detection using the Frozen Gaussian Approximation
• A deep learning based multiscale approach to segment cancer area in liver whole slide image
• Butterfly Effect: Bidirectional Control of Classification Performance by Small Additive Perturbation
• An Aggressive Genetic Programming Approach for Searching Neural Network Structure Under Computational Constraints
• Faster Optimization-Based Meta-Learning Adaptation Phase
• MOSQUITO-NET: A deep learning based CADx system for malaria diagnosis along with model interpretation using GradCam and class activation maps
• AI4AI: Quantitative Methods for Classifying Host Species from Avian Influenza DNA Sequence
• Generative Adversarial Mapping Networks
• Multi-task learning for joint weakly-supervised segmentation and aortic arch anomaly classification in fetal cardiac MRI
• Learning Channel Inter-dependencies at Multiple Scales on Dense Networks for Face Recognition
• Graph-Adaptive Pruning for Efficient Inference of Convolutional Neural Networks
• A Deep Learning Application for Psoriasis Detection
• A Singular Value Perspective on Model Robustness
• Improving Uncertainty Calibration of Deep Neural Networks via Truth Discovery and Geometric Optimization
• DiffusionNet: Accelerating the solution of Time-Dependent partial differential equations using deep learning
• Evaluating Convolutional Neural Networks for COVID-19 classification in chest X-ray images
• Improving training of deep neural networks via Singular Value Bounding
• Gradually Updated Neural Networks for Large-Scale Image Recognition
• Learning Less-Overlapping Representations
• Deep Geodesic Learning for Segmentation and Anatomical Landmarking
• 3D Face Reconstruction from Light Field Images: A Model-free Approach
• ResNetX: a more disordered and deeper network architecture
• Non-Parametric Transformation Networks
• Improving the Resolution of CNN Feature Maps Efficiently with Multisampling
• Exploiting Operation Importance for Differentiable Neural Architecture Search
• Investigating Transfer Learning Capabilities of Vision Transformers and CNNs by Fine-Tuning a Single Trainable Block
• 3DContextNet: K-d Tree Guided Hierarchical Learning of Point Clouds Using Local and Global Contextual Cues
• Dense xUnit Networks
• GM-Net: Learning Features with More Efficiency
• Learning Multimodal Fixed-Point Weights using Gradient Descent
• Feature-Fused Context-Encoding Network for Neuroanatomy Segmentation
• Exploring Weight Symmetry in Deep Neural Networks
• Large Batch Training Does Not Need Warmup
• Single Image Depth Estimation Trained via Depth from Defocus Cues
• Propagated Perturbation of Adversarial Attack for well-known CNNs: Empirical Study and its Explanation
• Wasserstein Routed Capsule Networks
• Deep Learning based Dimple Segmentation for Quantitative Fractography
• Extrapolation for Large-batch Training in Deep Learning
• Ensemble Transfer Learning for Emergency Landing Field Identification on Moderate Resource Heterogeneous Kubernetes Cluster
• Deep High-Resolution Network for Low Dose X-ray CT Denoising
• Model Optimization for Deep Space Exploration via Simulators and Deep Learning
• Custom Deep Neural Network for 3D Covid Chest CT-scan Classification
• Meta-Cal: Well-controlled Post-hoc Calibration by Ranking
• Empirical Analysis of Image Caption Generation using Deep Learning
• Learning to Embed Categorical Features without Embedding Tables for Recommendation
• An Artificial Intelligence-Driven Agent for Real-Time Head-and-Neck IMRT Plan Generation using Conditional Generative Adversarial Network (cGAN)
• Rotation Equivariant CNNs for Digital Pathology
• Learn Faster and Forget Slower via Fast and Stable Task Adaptation
• SuperNet -- An efficient method of neural networks ensembling
• Image Restoration Using Deep Regulated Convolutional Networks
• CrescendoNet: A Simple Deep Convolutional Neural Network with Ensemble Behavior
• Learning image from projection: a full-automatic reconstruction (FAR) net for sparse-views computed tomography
• WaveletNet: Logarithmic Scale Efficient Convolutional Neural Networks for Edge Devices
• NavigationNet: A Large-scale Interactive Indoor Navigation Dataset
• Deep Imbalanced Attribute Classification using Visual Attention Aggregation
• Deep Global-Connected Net With The Generalized Multi-Piecewise ReLU Activation in Deep Learning
• Tensor Switching Networks
• Semantic Segmentation via Highly Fused Convolutional Network with Multiple Soft Cost Functions
• Adversarial Policy Gradient for Deep Learning Image Augmentation
• ScieNet: Deep Learning with Spike-assisted Contextual Information Extraction
• Deep learning achieves perfect anomaly detection on 108,308 retinal images including unlearned diseases
• Multi-QuartzNet: Multi-Resolution Convolution for Speech Recognition with Multi-Layer Feature Fusion
• Neural Rejuvenation: Improving Deep Network Training by Enhancing Computational Resource Utilization
• Chair Segments: A Compact Benchmark for the Study of Object Segmentation
• On Batch Orthogonalization Layers
• ParaNet - Using Dense Blocks for Early Inference
• A comparison of deep machine learning algorithms in COVID-19 disease diagnosis
• Wasserstein Introspective Neural Networks
• Neuron Segmentation Using Deep Complete Bipartite Networks
• Approximate Fisher Information Matrix to Characterise the Training of Deep Neural Networks
• An Attention Model for group-level emotion recognition
• Focusing on Shadows for Predicting Heightmaps from Single Remotely Sensed RGB Images with Deep Learning
• NTIRE 2020 Challenge on Spectral Reconstruction from an RGB Image
• Filter Bank Regularization of Convolutional Neural Networks
• Augmentation Inside the Network
• Convolutional neural networks compression with low rank and sparse tensor decompositions
• Going Deeper with Lean Point Networks
• ADNet: Leveraging Error-Bias Towards Normal Direction in Face Alignment
• BPPSA: Scaling Back-propagation by Parallel Scan Algorithm
• Localized Compression: Applying Convolutional Neural Networks to Compressed Images
• Make Thunderbolts Less Frightening -- Predicting Extreme Weather Using Deep Learning
• Orthogonal and Idempotent Transformations for Learning Deep Neural Networks
• Towards Deeper Generative Architectures for GANs using Dense connections
• Real-time Light Estimation and Neural Soft Shadows for AR Indoor Scenarios
• CNN-based fully automatic mitral valve extraction using CT images and existence probability maps
• Synthesis of High-Quality Visible Faces from Polarimetric Thermal Faces using Generative Adversarial Networks
• DeepSIC: Deep Semantic Image Compression
• Manifestation of Image Contrast in Deep Networks
• Non-Local Representation based Mutual Affine-Transfer Network for Photorealistic Stylization
• High-Level Perceptual Similarity is Enabled by Learning Diverse Tasks
• ESFNet: Efficient Network for Building Extraction from High-Resolution Aerial Images
• Speeding-up Age Estimation in Intelligent Demographics System via Network Optimization
• Retrieval and Localization with Observation Constraints
• Effectiveness of Deep Networks in NLP using BiDAF as an example architecture
• signProx: One-Bit Proximal Algorithm for Nonconvex Stochastic Optimization
• VALUE: Large Scale Voting-based Automatic Labelling for Urban Environments
• Now that I can see, I can improve: Enabling data-driven finetuning of CNNs on the edge
• NTIRE 2020 Challenge on Video Quality Mapping: Methods and Results
• Deep Neural Networks with Short Circuits for Improved Gradient Learning
• Multi-Scale Weight Sharing Network for Image Recognition
• Towards More Efficient and Effective Inference: The Joint Decision of Multi-Participants
• Rethinking Fully Convolutional Networks for the Analysis of Photoluminescence Wafer Images
• TUNet: Incorporating segmentation maps to improve classification
• Deep Active Localization
• Informative sample generation using class aware generative adversarial networks for classification of chest Xrays
• Exploring Temporal Information for Improved Video Understanding
• The Compressed Model of Residual CNDS
• Attention Incorporate Network: A network can adapt various data size
• Latent Transformations for Object View Points Synthesis
• Exploiting Nontrivial Connectivity for Automatic Speech Recognition
• Automatically Designing CNN Architectures for Medical Image Segmentation
• Unbounded Output Networks for Classification
• Backtracking gradient descent method for general $C^1$ functions, with applications to Deep Learning
• Deep Room Recognition Using Inaudible Echos
• Reconciling Feature-Reuse and Overfitting in DenseNet with Specialized Dropout
• Penetrating the Fog: the Path to Efficient CNN Models
• Sequenced-Replacement Sampling for Deep Learning
• Channel selection using Gumbel Softmax
• Crowding in humans is unlike that in convolutional neural networks
• SwGridNet: A Deep Convolutional Neural Network based on Grid Topology for Image Classification
• Exposure Interpolation by Combining Model-driven and Data-driven Methods
• Adaptive Noise Injection: A Structure-Expanding Regularization for RNN
• Safe Augmentation: Learning Task-Specific Transformations from Data
• Left ventricle segmentation By modelling uncertainty in prediction of deep convolutional neural networks and adaptive thresholding inference
• Convergence Analysis of Gradient Descent Algorithms with Proportional Updates
• Deep Competitive Pathway Networks
• Block-term Tensor Neural Networks
• Convolutional-LSTM for Multi-Image to Single Output Medical Prediction
• When deep learning models on GPU can be accelerated by taking advantage of unstructured sparsity
• An Effective Training Method For Deep Convolutional Neural Network
• Multi-Glimpse Network: A Robust and Efficient Classification Architecture based on Recurrent Downsampled Attention
• Adaptive Transfer Learning: a simple but effective transfer learning
• Content-adaptive Representation Learning for Fast Image Super-resolution
• Adversarial Attacks with Time-Scale Representations
• Classification of COVID-19 from CXR Images in a 15-class Scenario: an Attempt to Avoid Bias in the System
• Scaling Laws for the Few-Shot Adaptation of Pre-trained Image Classifiers
• MARNet: Multi-Abstraction Refinement Network for 3D Point Cloud Analysis
• Deep Image Compositing
• SelectScale: Mining More Patterns from Images via Selective and Soft Dropout
• Deep Gradient Projection Networks for Pan-sharpening
• AET-EFN: A Versatile Design for Static and Dynamic Event-Based Vision
• ClaRe: Practical Class Incremental Learning By Remembering Previous Class Representations
• Performance Analysis of Out-of-Distribution Detection on Various Trained Neural Networks
• Reinforced Attention for Few-Shot Learning and Beyond
• Relative Depth Order Estimation Using Multi-scale Densely Connected Convolutional Networks