Pixel Recurrent Neural Networks

Abstract

Modeling the distribution of natural images is a landmark problem in unsupervised learning. This task requires an image model that is at once expressive, tractable and scalable. We present a deep neural network that sequentially predicts the pixels in an image along the two spatial dimensions. Our method models the discrete probability of the raw pixel values and encodes the complete set of dependencies in the image. Architectural novelties include fast two-dimensional recurrent layers and an effective use of residual connections in deep recurrent networks. We achieve log-likelihood scores on natural images that are considerably better than the previous state of the art. Our main results also provide benchmarks on the diverse ImageNet dataset. Samples generated from the model appear crisp, varied and globally coherent.

References in corpus (5)

• NICE: Non-linear Independent Components Estimation
• Markov Chain Monte Carlo and Variational Inference: Bridging the Gap
• MADE: Masked Autoencoder for Distribution Estimation
• Parallel Multi-Dimensional LSTM, With Application to Fast Biomedical Volumetric Image Segmentation
• Iterative Neural Autoregressive Distribution Estimator (NADE-k)

Cited by in corpus (431)

• Denoising Diffusion Probabilistic Models
• An Introduction to Variational Autoencoders
• Convolutional Sequence to Sequence Learning
• Conditional Image Synthesis With Auxiliary Classifier GANs
• Conditional Image Generation with PixelCNN Decoders
• XLNet: Generalized Autoregressive Pretraining for Language Understanding
• Evaluating Large Language Models Trained on Code
• End-to-end Optimized Image Compression
• Deep Learning for Single Image Super-Resolution: A Brief Review
• Language Modeling with Gated Convolutional Networks
• Generative Modeling by Estimating Gradients of the Data Distribution
• Toward Multimodal Image-to-Image Translation
• Bayesian Deep Convolutional Encoder-Decoder Networks for Surrogate Modeling and Uncertainty Quantification
• Density estimation using Real NVP
• Adversarially Learned Inference
• Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognition
• Recent Advances in Recurrent Neural Networks
• Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges
• This Looks Like That: Deep Learning for Interpretable Image Recognition
• Contrastive Multiview Coding
• Cascaded Diffusion Models for High Fidelity Image Generation
• Neural Audio Synthesis of Musical Notes with WaveNet Autoencoders
• CogView: Mastering Text-to-Image Generation via Transformers
• Adversarial Risk and the Dangers of Evaluating Against Weak Attacks
• How Generative Adversarial Networks and Their Variants Work: An Overview
• NVAE: A Deep Hierarchical Variational Autoencoder
• Axial Attention in Multidimensional Transformers
• Generating Multi-label Discrete Patient Records using Generative Adversarial Networks
• InfoVAE: Information Maximizing Variational Autoencoders
• Normalizing Flows for Probabilistic Modeling and Inference
• Count-Based Exploration with Neural Density Models
• Professor Forcing: A New Algorithm for Training Recurrent Networks
• Neural Image Compression via Non-Local Attention Optimization and Improved Context Modeling
• Recent Advances in Convolutional Neural Networks
• #Exploration: A Study of Count-Based Exploration for Deep Reinforcement Learning
• Non-local Attention Optimized Deep Image Compression
• A Review on Deep Learning Techniques for Video Prediction
• MetNet: A Neural Weather Model for Precipitation Forecasting
• Improved Precision and Recall Metric for Assessing Generative Models
• Differentiable Learning of Quantum Circuit Born Machine
• PixelDefend: Leveraging Generative Models to Understand and Defend against Adversarial Examples
• Variational Diffusion Models
• ChemTS: An Efficient Python Library for de novo Molecular Generation
• Unsupervised Generative Modeling Using Matrix Product States
• Auxiliary Deep Generative Models
• Identification of Deep Network Generated Images Using Disparities in Color Components
• Ladder Variational Autoencoders
• TensorFlow Distributions
• Unifying Count-Based Exploration and Intrinsic Motivation
• Learning What and Where to Draw
• GraphAF: a Flow-based Autoregressive Model for Molecular Graph Generation
• Stochastic Adversarial Video Prediction
• Adversarial Text-to-Image Synthesis: A Review
• Emergence of grid-like representations by training recurrent neural networks to perform spatial localization
• Masked Autoregressive Flow for Density Estimation
• QuaterNet: A Quaternion-based Recurrent Model for Human Motion
• Generative replay with feedback connections as a general strategy for continual learning
• Glow: Generative Flow with Invertible 1x1 Convolutions
• Improving Variational Inference with Inverse Autoregressive Flow
• Causal Contextual Prediction for Learned Image Compression
• Mining gold from implicit models to improve likelihood-free inference
• Invertible Residual Networks
• Deep Learning-Based Video Coding: A Review and A Case Study
• Design of metalloproteins and novel protein folds using variational autoencoders
• A Guide to Constraining Effective Field Theories with Machine Learning
• Neural Network Renormalization Group
• VideoGPT: Video Generation using VQ-VAE and Transformers
• Learning for Video Compression
• Deep Identity-aware Transfer of Facial Attributes
• Generative Probabilistic Novelty Detection with Adversarial Autoencoders
• Representations of language in a model of visually grounded speech signal
• Sequence-to-point learning with neural networks for nonintrusive load monitoring
• 3D Self-Supervised Methods for Medical Imaging
• Image coding for machines: an end-to-end learned approach
• BIVA: A Very Deep Hierarchy of Latent Variables for Generative Modeling
• Asymptotically unbiased estimation of physical observables with neural samplers
• Super-Resolution via Deep Learning
• Generating Diverse High-Fidelity Images with VQ-VAE-2
• Sample Efficient Adaptive Text-to-Speech
• Generative Models for Automatic Chemical Design
• Implicit Generation and Generalization in Energy-Based Models
• How to train your neural ODE: the world of Jacobian and kinetic regularization
• A Neural Dirichlet Process Mixture Model for Task-Free Continual Learning
• High-Fidelity Image Generation With Fewer Labels
• Learning and Evaluating Representations for Deep One-class Classification
• Learning and Inference on Generative Adversarial Quantum Circuits
• Generative Face Completion
• Image Synthesis with a Single (Robust) Classifier
• Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representations
• Variational Deep Embedding: An Unsupervised and Generative Approach to Clustering
• Structured Denoising Diffusion Models in Discrete State-Spaces
• How to Train Your Energy-Based Models
• ChatPainter: Improving Text to Image Generation using Dialogue
• Copyright in Generative Deep Learning
• Mining for Dark Matter Substructure: Inferring subhalo population properties from strong lenses with machine learning
• Learning Memory Access Patterns
• SATAR: A Self-supervised Approach to Twitter Account Representation Learning and its Application in Bot Detection
• PixelCNN++: Improving the PixelCNN with Discretized Logistic Mixture Likelihood and Other Modifications
• Towards an integration of deep learning and neuroscience
• Efficient Neural Audio Synthesis
• Deep Generative Modelling: A Comparative Review of VAEs, GANs, Normalizing Flows, Energy-Based and Autoregressive Models
• Sequential Neural Likelihood: Fast Likelihood-free Inference with Autoregressive Flows
• Deep Representation Learning in Speech Processing: Challenges, Recent Advances, and Future Trends
• PointFlow: 3D Point Cloud Generation with Continuous Normalizing Flows
• Contracting Arbitrary Tensor Networks: General Approximate Algorithm and Applications in Graphical Models and Quantum Circuit Simulations
• PolyGen: An Autoregressive Generative Model of 3D Meshes
• Discrete Sequential Prediction of Continuous Actions for Deep RL
• Overparameterized Neural Networks Implement Associative Memory
• Statistical post-processing of wind speed forecasts using convolutional neural networks
• PasteGAN: A Semi-Parametric Method to Generate Image from Scene Graph
• Security and Privacy Issues in Deep Learning
• Implicit Maximum Likelihood Estimation
• CVAE-GAN: Fine-Grained Image Generation through Asymmetric Training
• Boosting Monte Carlo simulations of spin glasses using autoregressive neural networks
• A Survey of Deep Learning for Scientific Discovery
• Pervasive Attention: 2D Convolutional Neural Networks for Sequence-to-Sequence Prediction
• Learning to Perform Physics Experiments via Deep Reinforcement Learning
• Data Augmentation for Object Detection via Progressive and Selective Instance-Switching
• Bit-Swap: Recursive Bits-Back Coding for Lossless Compression with Hierarchical Latent Variables
• Learning Likelihoods with Conditional Normalizing Flows
• Deep Learning for Plasma Tomography and Disruption Prediction from Bolometer Data
• Generating Multi-Agent Trajectories using Programmatic Weak Supervision
• Photographic Text-to-Image Synthesis with a Hierarchically-nested Adversarial Network
• Deep symbolic regression: Recovering mathematical expressions from data via risk-seeking policy gradients
• Parallel Multiscale Autoregressive Density Estimation
• Learning Beyond Human Expertise with Generative Models for Dental Restorations
• Maximum Likelihood Training of Score-Based Diffusion Models
• Stacked Generative Adversarial Networks
• A Survey of Optimization Methods from a Machine Learning Perspective
• Generative adversarial interpolative autoencoding: adversarial training on latent space interpolations encourage convex latent distributions
• Boosting Occluded Image Classification via Subspace Decomposition Based Estimation of Deep Features
• High-Resolution Mammogram Synthesis using Progressive Generative Adversarial Networks
• Face Hallucination with Finishing Touches
• Robust Training of Vector Quantized Bottleneck Models
• Simple, Distributed, and Accelerated Probabilistic Programming
• DVAE++: Discrete Variational Autoencoders with Overlapping Transformations
• Unified Probabilistic Deep Continual Learning through Generative Replay and Open Set Recognition
• IDF++: Analyzing and Improving Integer Discrete Flows for Lossless Compression
• Adversarial Training in Affective Computing and Sentiment Analysis: Recent Advances and Perspectives
• High-Quality Self-Supervised Deep Image Denoising
• Boundless: Generative Adversarial Networks for Image Extension
• Semi-Supervised Histology Classification using Deep Multiple Instance Learning and Contrastive Predictive Coding
• PeCo: Perceptual Codebook for BERT Pre-training of Vision Transformers
• Unconditional Audio Generation with Generative Adversarial Networks and Cycle Regularization
• Investigating Object Compositionality in Generative Adversarial Networks
• On the Evaluation of Conditional GANs
• Autoregressive Quantile Networks for Generative Modeling
• VideoFlow: A Conditional Flow-Based Model for Stochastic Video Generation
• Pedestrian Attribute Recognition: A Survey
• Auto-Encoding Total Correlation Explanation
• GLoMo: Unsupervisedly Learned Relational Graphs as Transferable Representations
• Lossless Image Compression through Super-Resolution
• Improved Contrastive Divergence Training of Energy Based Models
• Combiner: Full Attention Transformer with Sparse Computation Cost
• Neural Spline Flows
• Detecting Photoshopped Faces by Scripting Photoshop
• Generating 3D faces using Convolutional Mesh Autoencoders
• High-Resolution Image Inpainting using Multi-Scale Neural Patch Synthesis
• Are Generative Classifiers More Robust to Adversarial Attacks?
• Generating equilibrium molecules with deep neural networks
• Divide, Denoise, and Defend against Adversarial Attacks
• High-resolution Deep Convolutional Generative Adversarial Networks
• Time-Agnostic Prediction: Predicting Predictable Video Frames
• Perfect density models cannot guarantee anomaly detection
• Hierarchical Autoregressive Image Models with Auxiliary Decoders
• Photo-Realistic Video Prediction on Natural Videos of Largely Changing Frames
• Unsupervised Doodling and Painting with Improved SPIRAL
• X-LXMERT: Paint, Caption and Answer Questions with Multi-Modal Transformers
• Audio Description from Image by Modal Translation Network
• Augmented Normalizing Flows: Bridging the Gap Between Generative Flows and Latent Variable Models
• BlockCNN: A Deep Network for Artifact Removal and Image Compression
• Iterative Retraining of Quantum Spin Models Using Recurrent Neural Networks
• Semi-supervised Image Attribute Editing using Generative Adversarial Networks
• Benchmarking Bonus-Based Exploration Methods on the Arcade Learning Environment
• Gang of GANs: Generative Adversarial Networks with Maximum Margin Ranking
• Autoencoders for music sound modeling: a comparison of linear, shallow, deep, recurrent and variational models
• Calculating Renyi Entropies with Neural Autoregressive Quantum States
• Neural Approximate Sufficient Statistics for Implicit Models
• No Padding Please: Efficient Neural Handwriting Recognition
• Adversarial Learning for Improved Onsets and Frames Music Transcription
• Effective LHC measurements with matrix elements and machine learning
• SemanticAdv: Generating Adversarial Examples via Attribute-conditional Image Editing
• Pro-UIGAN: Progressive Face Hallucination from Occluded Thumbnails
• Disentangled Person Image Generation
• Video Compression through Image Interpolation
• Deep Perceptual Compression
• VFlow: More Expressive Generative Flows with Variational Data Augmentation
• Fast Generation for Convolutional Autoregressive Models
• Wavelet Flow: Fast Training of High Resolution Normalizing Flows
• Learning Convolutional Networks for Content-weighted Image Compression
• Learning Layout and Style Reconfigurable GANs for Controllable Image Synthesis
• Neural Autoregressive Distribution Estimation
• PointGrow: Autoregressively Learned Point Cloud Generation with Self-Attention
• Precomputed Real-Time Texture Synthesis with Markovian Generative Adversarial Networks
• Stochastic WaveNet: A Generative Latent Variable Model for Sequential Data
• Image-to-Image Translation: Methods and Applications
• Neural Attribute Machines for Program Generation
• Model Inversion Networks for Model-Based Optimization
• Likelihood-free inference with an improved cross-entropy estimator
• Transformation Autoregressive Networks
• Generative Latent Flow
• Hierarchical Adversarially Learned Inference
• Person Re-identification with Deep Similarity-Guided Graph Neural Network
• Unsupervised Person Image Synthesis in Arbitrary Poses
• Learning Discrete Distributions by Dequantization
• Continuous Melody Generation via Disentangled Short-Term Representations and Structural Conditions
• The Six Fronts of the Generative Adversarial Networks
• A Generative Model of Galactic Dust Emission Using Variational Inference
• Variational Autoencoders with Normalizing Flow Decoders
• Image Generation From Small Datasets via Batch Statistics Adaptation
• Ranking CGANs: Subjective Control over Semantic Image Attributes
• One Person, One Model, One World: Learning Continual User Representation without Forgetting
• FlipDial: A Generative Model for Two-Way Visual Dialogue
• Associative Compression Networks for Representation Learning
• Conditional Sum-Product Networks: Imposing Structure on Deep Probabilistic Architectures
• Policy Evaluation Networks
• DrumGAN: Synthesis of Drum Sounds With Timbral Feature Conditioning Using Generative Adversarial Networks
• Adversarial symmetric GANs: bridging adversarial samples and adversarial networks
• Scaling Autoregressive Video Models
• PixelVAE++: Improved PixelVAE with Discrete Prior
• TraDE: Transformers for Density Estimation
• Split-Brain Autoencoders: Unsupervised Learning by Cross-Channel Prediction
• Generating unseen complex scenes: are we there yet?
• Two Souls in an Adversarial Image: Towards Universal Adversarial Example Detection using Multi-view Inconsistency
• SANet: Structure-Aware Network for Visual Tracking
• Continual Learning in Recurrent Neural Networks
• Hidden Talents of the Variational Autoencoder
• Task Specific Adversarial Cost Function
• Preventing Posterior Collapse with delta-VAEs
• MAE: Mutual Posterior-Divergence Regularization for Variational AutoEncoders
• BINet: a binary inpainting network for deep patch-based image compression
• Generalization Properties of Optimal Transport GANs with Latent Distribution Learning
• Temporal Difference Variational Auto-Encoder
• Global-Local Bidirectional Reasoning for Unsupervised Representation Learning of 3D Point Clouds
• Road Segmentation Using CNN with GRU
• Hierarchical Patch VAE-GAN: Generating Diverse Videos from a Single Sample
• Noise2Void - Learning Denoising from Single Noisy Images
• Learning Deep Representations for Semantic Image Parsing: a Comprehensive Overview
• Densely connected normalizing flows
• Autoregressive Models: What Are They Good For?
• From Ansätze to Z-gates: a NASA View of Quantum Computing
• Learning to Collide: An Adaptive Safety-Critical Scenarios Generating Method
• Training Deep Energy-Based Models with f-Divergence Minimization
• Convolutional Normalizing Flows
• MintNet: Building Invertible Neural Networks with Masked Convolutions
• Improving Regression Performance with Distributional Losses
• Versatile Auxiliary Classifier with Generative Adversarial Network (VAC+GAN), Multi Class Scenarios
• COCO-GAN: Generation by Parts via Conditional Coordinating
• Invertible DenseNets with Concatenated LipSwish
• Integer Discrete Flows and Lossless Compression
• Disentangling neural mechanisms for perceptual grouping
• Unsupervised Anomaly Detection with Adversarial Mirrored AutoEncoders
• Generating Diverse Structure for Image Inpainting With Hierarchical VQ-VAE
• Coupling the reduced-order model and the generative model for an importance sampling estimator
• Learning deep kernels for exponential family densities
• Low-rank Characteristic Tensor Density Estimation Part II: Compression and Latent Density Estimation
• On the Road with 16 Neurons: Mental Imagery with Bio-inspired Deep Neural Networks
• Generating Images with Sparse Representations
• Learning Gradient Fields for Shape Generation
• A Contrastive Learning Approach for Training Variational Autoencoder Priors
• Conditioning Deep Generative Raw Audio Models for Structured Automatic Music
• Hierarchical VAEs Know What They Don't Know
• Numerically exact mimicking of quantum gas microscopy for interacting lattice fermions
• Sum-Product-Quotient Networks
• Likelihood Ratios for Out-of-Distribution Detection
• Neural Density Estimation and Likelihood-free Inference
• MAST: A Memory-Augmented Self-supervised Tracker
• Spatial Dependency Networks: Neural Layers for Improved Generative Image Modeling
• DUAL-GLOW: Conditional Flow-Based Generative Model for Modality Transfer
• Universal Value Density Estimation for Imitation Learning and Goal-Conditioned Reinforcement Learning
• Hybrid Physical-Deep Learning Model for Astronomical Inverse Problems
• Locally Masked Convolution for Autoregressive Models
• Soft then Hard: Rethinking the Quantization in Neural Image Compression
• Rethinking Recurrent Neural Networks and Other Improvements for Image Classification
• Compositional Visual Generation and Inference with Energy Based Models
• Predictive Sampling with Forecasting Autoregressive Models
• APRICOT: A Dataset of Physical Adversarial Attacks on Object Detection
• Image Generation from Layout
• Label-Noise Robust Generative Adversarial Networks
• Investigation of Using Disentangled and Interpretable Representations for One-shot Cross-lingual Voice Conversion
• Semantic Object Parsing with Graph LSTM
• Finding online neural update rules by learning to remember
• Training VAEs Under Structured Residuals
• Sequential Neural Methods for Likelihood-free Inference
• PixelCNN Models with Auxiliary Variables for Natural Image Modeling
• Detecting Out-of-distribution Samples via Variational Auto-encoder with Reliable Uncertainty Estimation
• Unleashing Transformers: Parallel Token Prediction with Discrete Absorbing Diffusion for Fast High-Resolution Image Generation from Vector-Quantized Codes
• Coupled Recurrent Models for Polyphonic Music Composition
• Amortized Bethe Free Energy Minimization for Learning MRFs
• Composing Music with Grammar Argumented Neural Networks and Note-Level Encoding
• Learning Content-Weighted Deep Image Compression
• Teaching Machines to Converse
• Biphasic Learning of GANs for High-Resolution Image-to-Image Translation
• L3C-Stereo: Lossless Compression for Stereo Images
• Deep Gaussian Markov Random Fields
• Benchmarking Deep Sequential Models on Volatility Predictions for Financial Time Series
• LUCSS: Language-based User-customized Colourization of Scene Sketches
• Hybrid Generative-Contrastive Representation Learning
• MaCow: Masked Convolutional Generative Flow
• Learning Context-Based Non-local Entropy Modeling for Image Compression
• Parametric context adaptive Laplace distribution for multimedia compression
• BachGAN: High-Resolution Image Synthesis from Salient Object Layout
• Cooperative Training of Fast Thinking Initializer and Slow Thinking Solver for Conditional Learning
• Generalized Energy Based Models
• A Simple Convolutional Generative Network for Next Item Recommendation
• Learning Implicit Generative Models by Teaching Explicit Ones
• Deterministic Decoding for Discrete Data in Variational Autoencoders
• Teaching Machines to Code: Neural Markup Generation with Visual Attention
• Generative Adversarial Forests for Better Conditioned Adversarial Learning
• NanoFlow: Scalable Normalizing Flows with Sublinear Parameter Complexity
• Variational Capsules for Image Analysis and Synthesis
• Class-Distinct and Class-Mutual Image Generation with GANs
• Parallel and Flexible Sampling from Autoregressive Models via Langevin Dynamics
• Filtering Variational Objectives
• On the Out-of-distribution Generalization of Probabilistic Image Modelling
• Deep Learned Frame Prediction for Video Compression
• Manifold-valued Image Generation with Wasserstein Generative Adversarial Nets
• Do Language Embeddings Capture Scales?
• Generative Image Modeling using Style and Structure Adversarial Networks
• Stochastic Conditional Generative Networks with Basis Decomposition
• To learn image super-resolution, use a GAN to learn how to do image degradation first
• Unpaired Learning of Deep Image Denoising
• Versatile Auxiliary Classifier with Generative Adversarial Network (VAC+GAN)
• Learning End-to-End Lossy Image Compression: A Benchmark
• Learning compact generalizable neural representations supporting perceptual grouping
• Backpropagation for Implicit Spectral Densities
• SwapText: Image Based Texts Transfer in Scenes
• Reconstruction of Simulation-Based Physical Field by Reconstruction Neural Network Method
• VAE-KRnet and its applications to variational Bayes
• Sensorimotor Visual Perception on Embodied System Using Free Energy Principle
• Efficient Learning of Generative Models via Finite-Difference Score Matching
• Spurious samples in deep generative models: bug or feature?
• Decentralized Multi-Agent Actor-Critic with Generative Inference
• Adversarial Robustness of Flow-Based Generative Models
• Neural Multi-scale Image Compression
• Neural Machine Translation: A Review and Survey
• How good is my GAN?
• Diverse feature visualizations reveal invariances in early layers of deep neural networks
• Gradient Origin Networks
• Fair Normalizing Flows
• Likelihood Contribution based Multi-scale Architecture for Generative Flows
• Curriculum Learning for Deep Generative Models with Clustering
• Semi-Implicit Generative Model
• MsCGAN: Multi-scale Conditional Generative Adversarial Networks for Person Image Generation
• Quantitatively rating galaxy simulations against real observations with anomaly detection
• Channel-Recurrent Autoencoding for Image Modeling
• Hyperparameter optimization with REINFORCE and Transformers
• Semantic Image Manipulation Using Scene Graphs
• Context-aware Padding for Semantic Segmentation
• Closing the Dequantization Gap: PixelCNN as a Single-Layer Flow
• Tractable Density Estimation on Learned Manifolds with Conformal Embedding Flows
• Controllable Image Synthesis via SegVAE
• Woodbury Transformations for Deep Generative Flows
• The ELBO of Variational Autoencoders Converges to a Sum of Three Entropies
• Learning to Structure an Image with Few Colors
• SentenceMIM: A Latent Variable Language Model
• Blur, Noise, and Compression Robust Generative Adversarial Networks
• A Forest from the Trees: Generation through Neighborhoods
• Physical Primitive Decomposition
• High Order Recurrent Neural Networks for Acoustic Modelling
• Scene Parsing via Dense Recurrent Neural Networks with Attentional Selection
• Re-examination of the Role of Latent Variables in Sequence Modeling
• A Tale of Three Probabilistic Families: Discriminative, Descriptive and Generative Models
• Handwriting styles: benchmarks and evaluation metrics
• RawNet: Fast End-to-End Neural Vocoder
• Text-to-image Synthesis via Symmetrical Distillation Networks
• Informative Sample Mining Network for Multi-Domain Image-to-Image Translation
• Generative Creativity: Adversarial Learning for Bionic Design
• Decoupling Global and Local Representations via Invertible Generative Flows
• Color Visual Illusions: A Statistics-based Computational Model
• EMIXER: End-to-end Multimodal X-ray Generation via Self-supervision
• Projected Latent Markov Chain Monte Carlo: Conditional Sampling of Normalizing Flows
• Calibrated Prediction Intervals for Neural Network Regressors
• Partition and Code: learning how to compress graphs
• Autoregressive Diffusion Models
• Learning Signal-Agnostic Manifolds of Neural Fields
• The Image Local Autoregressive Transformer
• Symmetric Wasserstein Autoencoders
• Learning of Colors from Color Names: Distribution and Point Estimation
• Deep Image Synthesis from Intuitive User Input: A Review and Perspectives
• ANFIC: Image Compression Using Augmented Normalizing Flows
• Causal affect prediction model using a facial image sequence
• Learning Neural Models for Natural Language Processing in the Face of Distributional Shift
• Multi Resolution LSTM For Long Term Prediction In Neural Activity Video
• End-to-End Image Compression with Probabilistic Decoding
• Learning Scalable ${\ell }_{\infty }$-constrained Near-lossless Image Compression via Joint Lossy Image and Residual Compression
• Normalizing Flows with Multi-Scale Autoregressive Priors
• One-Step Time-Dependent Future Video Frame Prediction with a Convolutional Encoder-Decoder Neural Network
• Towards Recurrent Autoregressive Flow Models
• Prototypical Recurrent Unit
• Recurrent Deconvolutional Generative Adversarial Networks with Application to Text Guided Video Generation
• Sibling Neural Estimators: Improving Iterative Image Decoding with Gradient Communication
• Deep Learning on Attributed Graphs: A Journey from Graphs to Their Embeddings and Back
• GAN-based Pose-aware Regulation for Video-based Person Re-identification
• Information Theoretic Lower Bounds on Negative Log Likelihood
• Actions Speak Louder than Listening: Evaluating Music Style Transfer based on Editing Experience
• End-To-End Dilated Variational Autoencoder with Bottleneck Discriminative Loss for Sound Morphing -- A Preliminary Study
• Latency-Controlled Neural Architecture Search for Streaming Speech Recognition
• Noise Robust Generative Adversarial Networks
• Lossless Compression with Latent Variable Models
• Learning Deep-Latent Hierarchies by Stacking Wasserstein Autoencoders
• Learned Image Compression with Soft Bit-based Rate-Distortion Optimization
• Generating new pictures in complex datasets with a simple neural network
• Discrete Variational Attention Models for Language Generation
• Generative Model for Heterogeneous Inference
• SINVAD: Search-based Image Space Navigation for DNN Image Classifier Test Input Generation
• Using an ensemble color space model to tackle adversarial examples
• Semi-Implicit Stochastic Recurrent Neural Networks
• GPCA: A Probabilistic Framework for Gaussian Process Embedded Channel Attention
• An Interpretable Generative Model for Handwritten Digit Image Synthesis
• Self-Reflective Variational Autoencoder
• ARMA Nets: Expanding Receptive Field for Dense Prediction
• Exploration into Translation-Equivariant Image Quantization
• Asymmetric Variational Autoencoders
• Face Super-Resolution Guided by 3D Facial Priors
• Quantised Transforming Auto-Encoders: Achieving Equivariance to Arbitrary Transformations in Deep Networks
• From Caesar Cipher to Unsupervised Learning: A New Method for Classifier Parameter Estimation
• Unsupervised Program Synthesis for Images By Sampling Without Replacement
• Progressive Spatial Recurrent Neural Network for Intra Prediction
• Neural Approximation of an Auto-Regressive Process through Confidence Guided Sampling
• Human Annotations Improve GAN Performances
• Layered Controllable Video Generation
• EdiBERT, a generative model for image editing
• A Transferable Adaptive Domain Adversarial Neural Network for Virtual Reality Augmented EMG-Based Gesture Recognition
• Probabilistic Autoencoder using Fisher Information
• Contributions to Large Scale Bayesian Inference and Adversarial Machine Learning
• Noise Contrastive Variational Autoencoders
• Hierarchical Reinforcement Learning with Optimal Level Synchronization Based on Flow-Based Deep Generative Model
• Natural Image Manipulation for Autoregressive Models Using Fisher Scores
• Knothe-Rosenblatt transport for Unsupervised Domain Adaptation
• PixelPyramids: Exact Inference Models from Lossless Image Pyramids