Self-Attention Generative Adversarial Networks

Abstract

In this paper, we propose the Self-Attention Generative Adversarial Network (SAGAN) which allows attention-driven, long-range dependency modeling for image generation tasks. Traditional convolutional GANs generate high-resolution details as a function of only spatially local points in lower-resolution feature maps. In SAGAN, details can be generated using cues from all feature locations. Moreover, the discriminator can check that highly detailed features in distant portions of the image are consistent with each other. Furthermore, recent work has shown that generator conditioning affects GAN performance. Leveraging this insight, we apply spectral normalization to the GAN generator and find that this improves training dynamics. The proposed SAGAN achieves the state-of-the-art results, boosting the best published Inception score from 36.8 to 52.52 and reducing Frechet Inception distance from 27.62 to 18.65 on the challenging ImageNet dataset. Visualization of the attention layers shows that the generator leverages neighborhoods that correspond to object shapes rather than local regions of fixed shape.

Cited by in corpus (505)

• Skillful Precipitation Nowcasting using Deep Generative Models of Radar
• Training Generative Adversarial Networks with Limited Data
• Alias-Free Generative Adversarial Networks
• DASNet: Dual attentive fully convolutional siamese networks for change detection of high resolution satellite images
• A General Survey on Attention Mechanisms in Deep Learning
• Attention in Natural Language Processing
• ResViT: Residual vision transformers for multi-modal medical image synthesis
• Deep Learning for Deepfakes Creation and Detection: A Survey
• ABCNet: Attentive Bilateral Contextual Network for Efficient Semantic Segmentation of Fine-Resolution Remote Sensing Images
• Deep Learning for Time Series Forecasting: Tutorial and Literature Survey
• PUFA-GAN: A Frequency-Aware Generative Adversarial Network for 3D Point Cloud Upsampling
• Sub-Image Anomaly Detection with Deep Pyramid Correspondences
• Escaping the Big Data Paradigm with Compact Transformers
• Improved Precision and Recall Metric for Assessing Generative Models
• Global Attention based Graph Convolutional Neural Networks for Improved Materials Property Prediction
• Generative Adversarial Networks in Computer Vision: A Survey and Taxonomy
• TransGAN: Two Pure Transformers Can Make One Strong GAN, and That Can Scale Up
• Semantic Image Synthesis with Spatially-Adaptive Normalization
• High-Performance Large-Scale Image Recognition Without Normalization
• An Overview of Deep Semi-Supervised Learning
• Vision Transformers For Weeds and Crops Classification Of High Resolution UAV Images
• UNETR: Transformers for 3D Medical Image Segmentation
• Dual Attention Network for Scene Segmentation
• An Explainable 3D Residual Self-Attention Deep Neural Network FOR Joint Atrophy Localization and Alzheimer's Disease Diagnosis using Structural MRI
• Restormer: Efficient Transformer for High-Resolution Image Restoration
• Transformers for Modeling Physical Systems
• Natural Adversarial Examples
• StyleNeRF: A Style-based 3D-Aware Generator for High-resolution Image Synthesis
• GAN Dissection: Visualizing and Understanding Generative Adversarial Networks
• Towards Unsupervised Deep Image Enhancement with Generative Adversarial Network
• Robust Attentional Aggregation of Deep Feature Sets for Multi-view 3D Reconstruction
• Scale- and Context-Aware Convolutional Non-intrusive Load Monitoring
• Adversarial Video Generation on Complex Datasets
• Selective Synthetic Augmentation with HistoGAN for Improved Histopathology Image Classification
• Exploring Spatial Significance via Hybrid Pyramidal Graph Network for Vehicle Re-identification
• PEPSI++: Fast and Lightweight Network for Image Inpainting
• Image Augmentations for GAN Training
• Train Sparsely, Generate Densely: Memory-efficient Unsupervised Training of High-resolution Temporal GAN
• Enhanced Balancing GAN: Minority-class Image Generation
• Review: Deep Learning in Electron Microscopy
• Jukebox: A Generative Model for Music
• High Fidelity Speech Synthesis with Adversarial Networks
• InSituNet: Deep Image Synthesis for Parameter Space Exploration of Ensemble Simulations
• InstaGAN: Instance-aware Image-to-Image Translation
• Differentiable Augmentation for Data-Efficient GAN Training
• ContraGAN: Contrastive Learning for Conditional Image Generation
• Noise Conscious Training of Non Local Neural Network powered by Self Attentive Spectral Normalized Markovian Patch GAN for Low Dose CT Denoising
• High-Fidelity Image Generation With Fewer Labels
• VTGAN: Semi-supervised Retinal Image Synthesis and Disease Prediction using Vision Transformers
• Vector-quantized Image Modeling with Improved VQGAN
• Image Synthesis with a Single (Robust) Classifier
• Efficient Attention: Attention with Linear Complexities
• Beyond Self-attention: External Attention using Two Linear Layers for Visual Tasks
• ProCAN: Progressive Growing Channel Attentive Non-Local Network for Lung Nodule Classification
• Few-Shot Unsupervised Image-to-Image Translation
• ViTGAN: Training GANs with Vision Transformers
• Copyright in Generative Deep Learning
• GODIVA: Generating Open-DomaIn Videos from nAtural Descriptions
• Deep Spatial Transformation for Pose-Guided Person Image Generation and Animation
• Asymmetric Non-local Neural Networks for Semantic Segmentation
• Unsupervised Cycle-consistent Generative Adversarial Networks for Pan-sharpening
• Parameter-Efficient Person Re-identification in the 3D Space
• IENet: Interacting Embranchment One Stage Anchor Free Detector for Orientation Aerial Object Detection
• The Universal Approximation Property
• Spatial-Angular Attention Network for Light Field Reconstruction
• Interpreting the Latent Space of GANs for Semantic Face Editing
• L2G Auto-encoder: Understanding Point Clouds by Local-to-Global Reconstruction with Hierarchical Self-Attention
• On Self Modulation for Generative Adversarial Networks
• Weak Supervision for Generating Pixel-Level Annotations in Scene Text Segmentation
• Deep Generative Modelling: A Comparative Review of VAEs, GANs, Normalizing Flows, Energy-Based and Autoregressive Models
• You Only Need Adversarial Supervision for Semantic Image Synthesis
• Learning quantum data with the quantum Earth Mover's distance
• Exploring the landscapes of "computing": digital, neuromorphic, unconventional -- and beyond
• SESAME: Semantic Editing of Scenes by Adding, Manipulating or Erasing Objects
• A Good Image Generator Is What You Need for High-Resolution Video Synthesis
• Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation
• CopyCat: Many-to-Many Fine-Grained Prosody Transfer for Neural Text-to-Speech
• Stabilizing Generative Adversarial Networks: A Survey
• SLSNet: Skin lesion segmentation using a lightweight generative adversarial network
• Attention-Enhanced Neural Network Models for Turbulence Simulation
• A Systematic Survey of Regularization and Normalization in GANs
• TileGAN: Synthesis of Large-Scale Non-Homogeneous Textures
• Unsupervised Object Segmentation by Redrawing
• PoNA: Pose-guided Non-local Attention for Human Pose Transfer
• CrossTransformers: spatially-aware few-shot transfer
• Space-time Mixing Attention for Video Transformer
• A Perspective on Deep Learning for Molecular Modeling and Simulations
• Reducing Noise in GAN Training with Variance Reduced Extragradient
• Detecting Cross-Modal Inconsistency to Defend Against Neural Fake News
• Improving Face-Based Age Estimation with Attention-Based Dynamic Patch Fusion
• Self-attending RNN for Speech Enhancement to Improve Cross-corpus Generalization
• On Leveraging Pretrained GANs for Generation with Limited Data
• LOGAN: Latent Optimisation for Generative Adversarial Networks
• Few-Shot Adaptation of Generative Adversarial Networks
• PSIGAN: Joint probabilistic segmentation and image distribution matching for unpaired cross-modality adaptation based MRI segmentation
• Normalization Techniques in Training DNNs: Methodology, Analysis and Application
• Is Attention Better Than Matrix Decomposition?
• RR-DnCNN v2.0: Enhanced Restoration-Reconstruction Deep Neural Network for Down-Sampling Based Video Coding
• LambdaNetworks: Modeling Long-Range Interactions Without Attention
• Is Generator Conditioning Causally Related to GAN Performance?
• Yes, we GAN: Applying Adversarial Techniques for Autonomous Driving
• Consistency Regularization for Generative Adversarial Networks
• Robust in Practice: Adversarial Attacks on Quantum Machine Learning
• AutoGAN: Neural Architecture Search for Generative Adversarial Networks
• CNN-generated images are surprisingly easy to spot... for now
• VCGAN: Video Colorization with Hybrid Generative Adversarial Network
• Creativity and Machine Learning: A Survey
• Rethinking Monocular Depth Estimation with Adversarial Training
• Attributing Fake Images to GANs: Learning and Analyzing GAN Fingerprints
• LFACon: Introducing Anglewise Attention to No-Reference Quality Assessment in Light Field Space
• A Review on Low-Dose Emission Tomography Post-Reconstruction Denoising with Neural Network Approaches
• CDSA: Cross-Dimensional Self-Attention for Multivariate, Geo-tagged Time Series Imputation
• Lower Dimensional Kernels for Video Discriminators
• STING: Self-attention based Time-series Imputation Networks using GAN
• Toward Interpretable Music Tagging with Self-Attention
• Quaternion Generative Adversarial Networks
• Spatial-temporal Conv-sequence Learning with Accident Encoding for Traffic Flow Prediction
• Cross-Modal Contrastive Learning for Text-to-Image Generation
• A Decentralized Parallel Algorithm for Training Generative Adversarial Nets
• Generative Adversarial Networks (GANs): An Overview of Theoretical Model, Evaluation Metrics, and Recent Developments
• DiverGAN: An Efficient and Effective Single-Stage Framework for Diverse Text-to-Image Generation
• A Large-Scale Study on Regularization and Normalization in GANs
• MelGAN-VC: Voice Conversion and Audio Style Transfer on arbitrarily long samples using Spectrograms
• Data-Efficient Instance Generation from Instance Discrimination
• PConv: Simple yet Effective Convolutional Layer for Generative Adversarial Network
• Language Guided Fashion Image Manipulation with Feature-wise Transformations
• Simple yet Effective Way for Improving the Performance of GAN
• GLoMo: Unsupervisedly Learned Relational Graphs as Transferable Representations
• Convolutional Generation of Textured 3D Meshes
• Attention2AngioGAN: Synthesizing Fluorescein Angiography from Retinal Fundus Images using Generative Adversarial Networks
• TinySpeech: Attention Condensers for Deep Speech Recognition Neural Networks on Edge Devices
• Adversarial score matching and improved sampling for image generation
• Focal Frequency Loss for Image Reconstruction and Synthesis
• The Lipschitz Constant of Self-Attention
• COVID-19 Screening Using Residual Attention Network an Artificial Intelligence Approach
• SkipConvGAN: Monaural Speech Dereverberation using Generative Adversarial Networks via Complex Time-Frequency Masking
• Unsupervised Adversarial Correction of Rigid MR Motion Artifacts
• CNN Attention Guidance for Improved Orthopedics Radiographic Fracture Classification
• Improved Consistency Regularization for GANs
• Normalized and Geometry-Aware Self-Attention Network for Image Captioning
• Deep Image Spatial Transformation for Person Image Generation
• Complement Face Forensic Detection and Localization with FacialLandmarks
• An Overview of Voice Conversion and its Challenges: From Statistical Modeling to Deep Learning
• Shape and Style GAN-based Multispectral Data Augmentation for Crop/Weed Segmentation in Precision Farming
• iQIYI-VID: A Large Dataset for Multi-modal Person Identification
• Involution: Inverting the Inherence of Convolution for Visual Recognition
• Progressive Sparse Local Attention for Video object detection
• Style Mixer: Semantic-aware Multi-Style Transfer Network
• Generative Adversarial Networks: A Survey Towards Private and Secure Applications
• Adversarially Learned Anomaly Detection
• Semi-supervised Image Attribute Editing using Generative Adversarial Networks
• DU-GAN: Generative Adversarial Networks with Dual-Domain U-Net Based Discriminators for Low-Dose CT Denoising
• ORDNet: Capturing Omni-Range Dependencies for Scene Parsing
• Controllable Person Image Synthesis with Attribute-Decomposed GAN
• TSSAT: Two-Stage Statistics-Aware Transformation for Artistic Style Transfer
• Subsampling Generative Adversarial Networks: Density Ratio Estimation in Feature Space with Softplus Loss
• Locality and compositionality in zero-shot learning
• A unified framework for 21cm tomography sample generation and parameter inference with Progressively Growing GANs
• Global Self-Attention Networks for Image Recognition
• Towards Better Understanding of Adaptive Gradient Algorithms in Generative Adversarial Nets
• Review of end-to-end speech synthesis technology based on deep learning
• Finding an Unsupervised Image Segmenter in Each of Your Deep Generative Models
• Self-Attention Capsule Networks for Object Classification
• Pixel-Adaptive Convolutional Neural Networks
• Towards Visually Explaining Variational Autoencoders
• Voice and accompaniment separation in music using self-attention convolutional neural network
• DSDANet: Deep Siamese Domain Adaptation Convolutional Neural Network for Cross-domain Change Detection
• PointGrow: Autoregressively Learned Point Cloud Generation with Self-Attention
• Exploring DeshuffleGANs in Self-Supervised Generative Adversarial Networks
• Source Generator Attribution via Inversion
• Connecting optical morphology, environment, and HI mass fraction for low-redshift galaxies using deep learning
• Disentangled High Quality Salient Object Detection
• Generative Adversarial Transformers
• HEI: hybrid explicit-implicit learning for multiscale problems
• Learning to Focus: Cascaded Feature Matching Network for Few-shot Image Recognition
• Reconstruction Student with Attention for Student-Teacher Pyramid Matching
• Crossing You in Style: Cross-modal Style Transfer from Music to Visual Arts
• TorchGAN: A Flexible Framework for GAN Training and Evaluation
• GTA: Global Temporal Attention for Video Action Understanding
• CycleGAN-based Non-parallel Speech Enhancement with an Adaptive Attention-in-attention Mechanism
• Attention-Guided Generative Adversarial Networks for Unsupervised Image-to-Image Translation
• A Review of Learning with Deep Generative Models from Perspective of Graphical Modeling
• GS-Net: Global Self-Attention Guided CNN for Multi-Stage Glaucoma Classification
• Top-k Training of GANs: Improving GAN Performance by Throwing Away Bad Samples
• Revisiting Stochastic Extragradient
• Multi-Path Region Mining For Weakly Supervised 3D Semantic Segmentation on Point Clouds
• The Six Fronts of the Generative Adversarial Networks
• Image Generation From Small Datasets via Batch Statistics Adaptation
• SALSA-TEXT : self attentive latent space based adversarial text generation
• Robust End-to-End Focal Liver Lesion Detection using Unregistered Multiphase Computed Tomography Images
• Differentially Private Generative Adversarial Networks with Model Inversion
• SOE-Net: A Self-Attention and Orientation Encoding Network for Point Cloud based Place Recognition
• Local Class-Specific and Global Image-Level Generative Adversarial Networks for Semantic-Guided Scene Generation
• AIM 2020: Scene Relighting and Illumination Estimation Challenge
• Triple consistency loss for pairing distributions in GAN-based face synthesis
• PSA-GAN: Progressive Self Attention GANs for Synthetic Time Series
• Convolution-enhanced Evolving Attention Networks
• Transformation-based Adversarial Video Prediction on Large-Scale Data
• Improving the Speed and Quality of GAN by Adversarial Training
• Planning on the fast lane: Learning to interact using attention mechanisms in path integral inverse reinforcement learning
• How Will It Drape Like? Capturing Fabric Mechanics from Depth Images
• Crowd-Robot Interaction: Crowd-aware Robot Navigation with Attention-based Deep Reinforcement Learning
• PU-GAN: a Point Cloud Upsampling Adversarial Network
• Parallel Scheduled Sampling
• SIGAN: A Novel Image Generation Method for Solar Cell Defect Segmentation and Augmentation
• Studying the potential of Graphcore IPUs for applications in Particle Physics
• MXR-U-Nets for Real Time Hyperspectral Reconstruction
• Regularizing Generative Adversarial Networks under Limited Data
• Contextual colorization and denoising for low-light ultra high resolution sequences
• Generative Convolution Layer for Image Generation
• TransCT: Dual-path Transformer for Low Dose Computed Tomography
• Compositional Transformers for Scene Generation
• cGANs with Multi-Hinge Loss
• Reverse Engineering of Generative Models: Inferring Model Hyperparameters from Generated Images
• SoK: Machine Learning Governance
• Normalized Attention Without Probability Cage
• Automated Olfactory Bulb Segmentation on High Resolutional T2-Weighted MRI
• FaR-GAN for One-Shot Face Reenactment
• PolyGAN: High-Order Polynomial Generators
• Order and Chaos: NTK views on DNN Normalization, Checkerboard and Boundary Artifacts
• Attentive Adversarial Learning for Domain-Invariant Training
• MotionVideoGAN: A Novel Video Generator Based on the Motion Space Learned from Image Pairs
• FREA-Unet: Frequency-aware U-net for Modality Transfer
• GraN-GAN: Piecewise Gradient Normalization for Generative Adversarial Networks
• A Deep Generative Model for Reordering Adjacency Matrices
• Realistic Ultrasound Image Synthesis for Improved Classification of Liver Disease
• A spin-glass model for the loss surfaces of generative adversarial networks
• Learning disconnected manifolds: a no GANs land
• COCO-GAN: Generation by Parts via Conditional Coordinating
• GATCluster: Self-Supervised Gaussian-Attention Network for Image Clustering
• Zero-Shot Learning from scratch (ZFS): leveraging local compositional representations
• Task-Aware Feature Generation for Zero-Shot Compositional Learning
• AlphaGAN: Fully Differentiable Architecture Search for Generative Adversarial Networks
• Adaptive Generation of Unrestricted Adversarial Inputs
• Twin Auxiliary Classifiers GAN
• Super Resolution Using Segmentation-Prior Self-Attention Generative Adversarial Network
• 6D Pose Estimation with Correlation Fusion
• In-Domain GAN Inversion for Real Image Editing
• Semantic Bottleneck Scene Generation
• FDFtNet: Facing Off Fake Images using Fake Detection Fine-tuning Network
• WaveFill: A Wavelet-based Generation Network for Image Inpainting
• Group Equivariant Generative Adversarial Networks
• Anycost GANs for Interactive Image Synthesis and Editing
• Understanding the Effectiveness of Lipschitz-Continuity in Generative Adversarial Nets
• A Multi-attribute Controllable Generative Model for Histopathology Image Synthesis
• Training Generative Reversible Networks
• Spatial Dependency Networks: Neural Layers for Improved Generative Image Modeling
• An Attention-Based System for Damage Assessment Using Satellite Imagery
• GDPP: Learning Diverse Generations Using Determinantal Point Process
• Self-Supervised GAN to Counter Forgetting
• Continuous Conditional Generative Adversarial Networks: Novel Empirical Losses and Label Input Mechanisms
• Conditional GANs with Auxiliary Discriminative Classifier
• Complex-Valued Time-Frequency Self-Attention for Speech Dereverberation
• Line Segment Detection Using Transformers without Edges
• Scale-Aware Network with Regional and Semantic Attentions for Crowd Counting under Cluttered Background
• Efficient Coarse-to-Fine Non-Local Module for the Detection of Small Objects
• Few-shot Video-to-Video Synthesis
• Dense CNN with Self-Attention for Time-Domain Speech Enhancement
• Self-Supervised GANs with Label Augmentation
• Multiple Instance Segmentation in Brachial Plexus Ultrasound Image Using BPMSegNet
• Data-Efficient GAN Training Beyond (Just) Augmentations: A Lottery Ticket Perspective
• Teachers Do More Than Teach: Compressing Image-to-Image Models
• Self Sparse Generative Adversarial Networks
• Hijack-GAN: Unintended-Use of Pretrained, Black-Box GANs
• Label-Noise Robust Generative Adversarial Networks
• RetinaGAN: An Object-aware Approach to Sim-to-Real Transfer
• Learning the Precise Feature for Cluster Assignment
• Parallax Attention for Unsupervised Stereo Correspondence Learning
• CoCoFuzzing: Testing Neural Code Models with Coverage-Guided Fuzzing
• Self-Supervised GANs via Auxiliary Rotation Loss
• A Decentralized Adaptive Momentum Method for Solving a Class of Min-Max Optimization Problems
• Toward Accurate and Realistic Virtual Try-on Through Shape Matching and Multiple Warps
• Image Synthesis From Reconfigurable Layout and Style
• MineGAN: effective knowledge transfer from GANs to target domains with few images
• RL-GAN-Net: A Reinforcement Learning Agent Controlled GAN Network for Real-Time Point Cloud Shape Completion
• PE-GAN: Prior Embedding GAN for PXD images at Belle II
• Discriminator Feature-based Inference by Recycling the Discriminator of GANs
• Headset: Human emotion awareness under partial occlusions multimodal dataset
• cGANs with Conditional Convolution Layer
• TSIT: A Simple and Versatile Framework for Image-to-Image Translation
• Enjoy Your Editing: Controllable GANs for Image Editing via Latent Space Navigation
• Decorating Your Own Bedroom: Locally Controlling Image Generation with Generative Adversarial Networks
• CaEGCN: Cross-Attention Fusion based Enhanced Graph Convolutional Network for Clustering
• Latent Normalizing Flows for Many-to-Many Cross-Domain Mappings
• Towards Deep and Efficient: A Deep Siamese Self-Attention Fully Efficient Convolutional Network for Change Detection in VHR Images
• Attention Distillation for Learning Video Representations
• CPGAN: Full-Spectrum Content-Parsing Generative Adversarial Networks for Text-to-Image Synthesis
• TuiGAN: Learning Versatile Image-to-Image Translation with Two Unpaired Images
• Dual-Attention GAN for Large-Pose Face Frontalization
• CoCosNet v2: Full-Resolution Correspondence Learning for Image Translation
• Few-shot Learning for Domain-specific Fine-grained Image Classification
• Telescoping Density-Ratio Estimation
• Multi-Head Linear Attention Generative Adversarial Network for Thin Cloud Removal
• Attentive Normalization for Conditional Image Generation
• A Deep Attentive Convolutional Neural Network for Automatic Cortical Plate Segmentation in Fetal MRI
• High Resolution Face Editing with Masked GAN Latent Code Optimization
• Bidirectional Generative Modeling Using Adversarial Gradient Estimation
• Scalable Balanced Training of Conditional Generative Adversarial Neural Networks on Image Data
• Colorization Transformer
• Complete the Look: Scene-based Complementary Product Recommendation
• Dual Contrastive Loss and Attention for GANs
• MP3net: coherent, minute-long music generation from raw audio with a simple convolutional GAN
• Spatial-Temporal Super-Resolution of Satellite Imagery via Conditional Pixel Synthesis
• On Characterizing GAN Convergence Through Proximal Duality Gap
• GANs Can Play Lottery Tickets Too
• An Acceleration Framework for High Resolution Image Synthesis
• Model Imitation for Model-Based Reinforcement Learning
• Generative Zero-shot Network Quantization
• Multiresolution Transformer Networks: Recurrence is Not Essential for Modeling Hierarchical Structure
• HumanGAN: A Generative Model of Humans Images
• GIQA: Generated Image Quality Assessment
• Bridging Global Context Interactions for High-Fidelity Image Completion
• Score-Guided Generative Adversarial Networks
• Online Knowledge Distillation with Diverse Peers
• TrustMAE: A Noise-Resilient Defect Classification Framework using Memory-Augmented Auto-Encoders with Trust Regions
• Fibro-CoSANet: Pulmonary Fibrosis Prognosis Prediction using a Convolutional Self Attention Network
• Adaptive Video Highlight Detection by Learning from User History
• Attention-Guided Generative Adversarial Network for Whisper to Normal Speech Conversion
• Learning Algorithms for Coarsening Uncertainty Space and Applications to Multiscale Simulations
• Learning Oracle Attention for High-fidelity Face Completion
• Context-Aware Domain Adaptation in Semantic Segmentation
• Variational Conditional GAN for Fine-grained Controllable Image Generation
• DRANet: Disentangling Representation and Adaptation Networks for Unsupervised Cross-Domain Adaptation
• Online Knowledge Distillation via Multi-branch Diversity Enhancement
• MUST-GAN: Multi-level Statistics Transfer for Self-driven Person Image Generation
• GuidedStyle: Attribute Knowledge Guided Style Manipulation for Semantic Face Editing
• Audio-Visual Event Localization via Recursive Fusion by Joint Co-Attention
• A Large-Scale, Time-Synchronized Visible and Thermal Face Dataset
• Semantically Multi-modal Image Synthesis
• Selective Synthetic Augmentation with Quality Assurance
• PSGAN: A Generative Adversarial Network for Remote Sensing Image Pan-Sharpening
• Attentive Representation Learning with Adversarial Training for Short Text Clustering
• World-Consistent Video-to-Video Synthesis
• Improved Transformer for High-Resolution GANs
• Ear2Face: Deep Biometric Modality Mapping
• Deep Magnification-Flexible Upsampling over 3D Point Clouds
• Generative Adversarial Network based Heuristics for Sampling-based Path Planning
• NSCaching: Simple and Efficient Negative Sampling for Knowledge Graph Embedding
• New Ideas and Trends in Deep Multimodal Content Understanding: A Review
• Point Cloud Upsampling via Disentangled Refinement
• DTGAN: Dual Attention Generative Adversarial Networks for Text-to-Image Generation
• Learning Disentangled Representations with Latent Variation Predictability
• BalaGAN: Image Translation Between Imbalanced Domains via Cross-Modal Transfer
• SSCR: Iterative Language-Based Image Editing via Self-Supervised Counterfactual Reasoning
• Multimodal Image-to-Image Translation via Mutual Information Estimation and Maximization
• Class-Distinct and Class-Mutual Image Generation with GANs
• Self-Supervised Ultrasound to MRI Fetal Brain Image Synthesis
• Sentence Guided Temporal Modulation for Dynamic Video Thumbnail Generation
• SDA-GAN: Unsupervised Image Translation Using Spectral Domain Attention-Guided Generative Adversarial Network
• Learning to Segment Brain Anatomy from 2D Ultrasound with Less Data
• Entropy-regularized Optimal Transport Generative Models
• SwapText: Image Based Texts Transfer in Scenes
• How Old Are You? Face Age Translation with Identity Preservation Using GANs
• Exploring Reciprocal Attention for Salient Object Detection by Cooperative Learning
• Estimation with Uncertainty via Conditional Generative Adversarial Networks
• Reinforcement Learning-based Visual Navigation with Information-Theoretic Regularization
• Expression-aware video inpainting for HMD removal in XR applications
• Full-attention based Neural Architecture Search using Context Auto-regression
• Improving Augmentation and Evaluation Schemes for Semantic Image Synthesis
• Slimmable Generative Adversarial Networks
• Towards DeepSentinel: An extensible corpus of labelled Sentinel-1 and -2 imagery and a general-purpose sensor-fusion semantic embedding model
• Graph-Based Continual Learning
• Independent Encoder for Deep Hierarchical Unsupervised Image-to-Image Translation
• Attend And Discriminate: Beyond the State-of-the-Art for Human Activity Recognition using Wearable Sensors
• FineFool: Fine Object Contour Attack via Attention
• Few-shot Compositional Font Generation with Dual Memory
• Assessing Deep Neural Networks as Probability Estimators
• RADDet: Range-Azimuth-Doppler based Radar Object Detection for Dynamic Road Users
• Quantitative Parametric Mapping of Tissues Properties from Standard Magnetic Resonance Imaging Enabled by Deep Learning
• Blur, Noise, and Compression Robust Generative Adversarial Networks
• Adversarial Joint Training with Self-Attention Mechanism for Robust End-to-End Speech Recognition
• Alleviation of Gradient Exploding in GANs: Fake Can Be Real
• TinyGAN: Distilling BigGAN for Conditional Image Generation
• Evolving Multi-Resolution Pooling CNN for Monaural Singing Voice Separation
• Efficient Heuristic Generation for Robot Path Planning with Recurrent Generative Model
• Zero-Shot Video Object Segmentation via Attentive Graph Neural Networks
• SMYRF: Efficient Attention using Asymmetric Clustering
• A Novel Framework for Selection of GANs for an Application
• Show, Attend and Translate: Unpaired Multi-Domain Image-to-Image Translation with Visual Attention
• Neural Function Modules with Sparse Arguments: A Dynamic Approach to Integrating Information across Layers
• Generative Adversarial Networks and Adversarial Autoencoders: Tutorial and Survey
• Collaborative Sampling in Generative Adversarial Networks
• Controllable Orthogonalization in Training DNNs
• A Comprehensive Approach to Unsupervised Embedding Learning based on AND Algorithm
• Generative-Discriminative Complementary Learning
• CLASS: Cross-Level Attention and Supervision for Salient Objects Detection
• An Empirical Study of Batch Normalization and Group Normalization in Conditional Computation
• Two-Stream Video Classification with Cross-Modality Attention
• Ensembling with Deep Generative Views
• Ensembles of GANs for synthetic training data generation
• Instance Selection for GANs
• not-so-BigGAN: Generating High-Fidelity Images on Small Compute with Wavelet-based Super-Resolution
• Towards Real-Time Advancement of Underwater Visual Quality with GAN
• Coarse Temporal Attention Network (CTA-Net) for Driver's Activity Recognition
• Scale Aware Adaptation for Land-Cover Classification in Remote Sensing Imagery
• Attributes Aware Face Generation with Generative Adversarial Networks
• Image Synthesis via Semantic Composition
• Realistic Image Synthesis with Configurable 3D Scene Layouts
• Learning who is in the market from time series: market participant discovery through adversarial calibration of multi-agent simulators
• Unrealistic Feature Suppression for Generative Adversarial Networks
• Deep Image Synthesis from Intuitive User Input: A Review and Perspectives
• Generative Modeling for Multi-task Visual Learning
• Toward Accurate and Realistic Outfits Visualization with Attention to Details
• Unsupervised Disentanglement of Linear-Encoded Facial Semantics
• Discover the Mysteries of the Maya: Selected Contributions from the Machine Learning Challenge & The Discovery Challenge Workshop at ECML PKDD 2021
• Momentum Centering and Asynchronous Update for Adaptive Gradient Methods
• Weakly-supervised Generative Adversarial Networks for medical image classification
• SpineOne: A One-Stage Detection Framework for Degenerative Discs and Vertebrae
• Collaborative Semantic Aggregation and Calibration for Federated Domain Generalization
• Stabilizing Training of Generative Adversarial Nets via Langevin Stein Variational Gradient Descent
• Network Architecture Search for Face Enhancement
• Noise Homogenization via Multi-Channel Wavelet Filtering for High-Fidelity Sample Generation in GANs
• Generative Adversarial Networks for photo to Hayao Miyazaki style cartoons
• Attribute-Modulated Generative Meta Learning for Zero-Shot Classification
• Video Playback Rate Perception for Self-supervisedSpatio-Temporal Representation Learning
• Exploiting Chain Rule and Bayes' Theorem to Compare Probability Distributions
• Deeper or Wider Networks of Point Clouds with Self-attention?
• S2cGAN: Semi-Supervised Training of Conditional GANs with Fewer Labels
• Attention-Guided Network for Iris Presentation Attack Detection
• Cross-View Image Synthesis with Deformable Convolution and Attention Mechanism
• Random Network Distillation as a Diversity Metric for Both Image and Text Generation
• Conditioning Trick for Training Stable GANs
• Generative networks as inverse problems with fractional wavelet scattering networks
• From Rain Generation to Rain Removal
• Imitating Unknown Policies via Exploration
• Feedback Attention for Cell Image Segmentation
• EMIXER: End-to-end Multimodal X-ray Generation via Self-supervision
• Structure-Aware Human-Action Generation
• Exploiting Non-Local Priors via Self-Convolution For Highly-Efficient Image Restoration
• Person-in-Context Synthesiswith Compositional Structural Space
• Kernel Self-Attention in Deep Multiple Instance Learning
• Neural Crossbreed: Neural Based Image Metamorphosis
• Generative Model without Prior Distribution Matching
• Making Method of Moments Great Again? -- How can GANs learn distributions
• A Novel Measure to Evaluate Generative Adversarial Networks Based on Direct Analysis of Generated Images
• Fine-grained Image-to-Image Transformation towards Visual Recognition
• On the Anomalous Generalization of GANs
• Exploiting Style and Attention in Real-World Super-Resolution
• G3AN: Disentangling Appearance and Motion for Video Generation
• SAG-VAE: End-to-end Joint Inference of Data Representations and Feature Relations
• UPI-Net: Semantic Contour Detection in Placental Ultrasound
• AtLoc: Attention Guided Camera Localization
• Multi-Frame Content Integration with a Spatio-Temporal Attention Mechanism for Person Video Motion Transfer
• Detecting GAN generated errors
• Hide-and-Tell: Learning to Bridge Photo Streams for Visual Storytelling
• Extractive Adversarial Networks: High-Recall Explanations for Identifying Personal Attacks in Social Media Posts
• Self-Attention Recurrent Network for Saliency Detection
• Interpretable Encrypted Searchable Neural Networks
• A deep learning framework for morphologic detail beyond the diffraction limit in infrared spectroscopic imaging
• HAD-GAN: A Human-perception Auxiliary Defense GAN to Defend Adversarial Examples
• Exploiting video sequences for unsupervised disentangling in generative adversarial networks
• Noise Robust Generative Adversarial Networks
• Irregular Convolutional Auto-Encoder on Point Clouds
• STC-Flow: Spatio-temporal Context-aware Optical Flow Estimation
• Do Not Forget to Attend to Uncertainty while Mitigating Catastrophic Forgetting
• When Relation Networks meet GANs: Relation GANs with Triplet Loss
• Realistic Face Reenactment via Self-Supervised Disentangling of Identity and Pose
• FusedProp: Towards Efficient Training of Generative Adversarial Networks
• Towards GANs' Approximation Ability
• An In-depth Walkthrough on Evolution of Neural Machine Translation
• Perceiving Music Quality with GANs
• Actor-Action Video Classification CSC 249/449 Spring 2020 Challenge Report
• Toward Zero-Shot Unsupervised Image-to-Image Translation
• Embedded Blockchains: A Synthesis of Blockchains, Spread Spectrum Watermarking, Perceptual Hashing & Digital Signatures
• Frame Aggregation and Multi-Modal Fusion Framework for Video-Based Person Recognition
• Difference-in-Differences: Bridging Normalization and Disentanglement in PG-GAN
• On the Global Self-attention Mechanism for Graph Convolutional Networks
• DeepAtrophy: Teaching a Neural Network to Differentiate Progressive Changes from Noise on Longitudinal MRI in Alzheimer's Disease
• Volumetric Transformer Networks
• Channel Pruning Guided by Spatial and Channel Attention for DNNs in Intelligent Edge Computing
• MixMix: All You Need for Data-Free Compression Are Feature and Data Mixing
• Unsupervised Image Transformation Learning via Generative Adversarial Networks
• Shape-Oriented Convolution Neural Network for Point Cloud Analysis
• Dual Projection Generative Adversarial Networks for Conditional Image Generation
• Blind Image Decomposition
• Ambient Hidden Space of Generative Adversarial Networks
• Incremental Scene Synthesis
• Domain-Specific Bias Filtering for Single Labeled Domain Generalization
• Re-designing cities with conditional adversarial networks
• Improving Differentiable Architecture Search with a Generative Model
• Learning from Mistakes based on Class Weighting with Application to Neural Architecture Search
• Synthesizing Photorealistic Images with Deep Generative Learning
• Linear Semantics in Generative Adversarial Networks
• Feature Completion for Occluded Person Re-Identification
• DINO: A Conditional Energy-Based GAN for Domain Translation
• ATP-Net: An Attention-based Ternary Projection Network For Compressed Sensing
• CanvasGAN: A simple baseline for text to image generation by incrementally patching a canvas
• Generative Adversarial Learning via Kernel Density Discrimination
• Multi-Channel Auto-Encoders and a Novel Dataset for Learning Domain Invariant Representations of Histopathology Images
• A Generative Adversarial Framework for Optimizing Image Matting and Harmonization Simultaneously
• Transferring Knowledge with Attention Distillation for Multi-Domain Image-to-Image Translation
• Can the Transformer Be Used as a Drop-in Replacement for RNNs in Text-Generating GANs?
• IAFA: Instance-aware Feature Aggregation for 3D Object Detection from a Single Image
• Instance-Conditioned GAN
• Online Refinement of Low-level Feature Based Activation Map for Weakly Supervised Object Localization
• Head2HeadFS: Video-based Head Reenactment with Few-shot Learning
• EvoGAN: An Evolutionary Computation Assisted GAN
• Perception Matters: Exploring Imperceptible and Transferable Anti-forensics for GAN-generated Fake Face Imagery Detection
• TB-Net: A Three-Stream Boundary-Aware Network for Fine-Grained Pavement Disease Segmentation
• A New Framework for Machine Intelligence: Concepts and Prototype
• Two-Stream Appearance Transfer Network for Person Image Generation
• CleftNet: Augmented Deep Learning for Synaptic Cleft Detection from Brain Electron Microscopy
• Relational Mimic for Visual Adversarial Imitation Learning
• Neural Network based Explicit Mixture Models and Expectation-maximization based Learning
• Modulated Self-attention Convolutional Network for VQA