Consistency Models
Explore this paper's citation graph
Summary
Consistency models are proposed, a new family of models that generate high quality samples by directly mapping noise to data that can outperform existing one-step, non-adversarial generative models on standard benchmarks such as CIFAR-10, ImageNet 64x64 and LSUN 256x256.
- Type
- preprint
- Published
- 2023-03-02
- Cited by
- 2,139
- References
- 88
- Access
- Open access
- OpenAlex
- https://openalex.org/W4323076585
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:257280191
Keywords
Computer science, Consistency (knowledge bases), Inpainting, Sampling (signal processing), Generative grammar
References
- LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loop
- Estimation of Non-Normalized Statistical Models by Score Matching
- Playing Atari with Deep Reinforcement Learning
- Stochastic Backpropagation and Approximate Inference in Deep Generative Models
- A Connection Between Score Matching and Denoising Autoencoders
- Modeling High-Dimensional Discrete Data with Multi-Layer Neural Networks
- ImageNet: A large-scale hierarchical image database
- Human-level control through deep reinforcement learning
- Pixel Recurrent Neural Networks
- Improved Techniques for Training GANs
- Self-Supervised GANs via Auxiliary Rotation Loss
- Improved Precision and Recall Metric for Assessing Generative Models
- Sliced Score Matching: A Scalable Approach to Density and Score Estimation
- Residual Flows for Invertible Generative Modeling
- Large Scale GAN Training for High Fidelity Natural Image Synthesis
- Generative Modeling by Estimating Gradients of the Data Distribution
- The Unreasonable Effectiveness of Deep Features as a Perceptual Metric
- Sliced Wasserstein Generative Models
- AutoGAN: Neural Architecture Search for Generative Adversarial Networks
- Analyzing and Improving the Image Quality of StyleGAN
Cited by
- Cloud storage availability and performance assessment: a study based on NoSQL DBMS
- The Logical Timestamp Skew Anomaly in Event-Replicated Transaction Schedulers
- Sequential Neural Score Estimation: Likelihood-Free Inference with Conditional Score Based Diffusion Models
- Boomerang: Local sampling on image manifolds using diffusion models
- Diffusion Denoising Process for Perceptron Bias in Out-of-distribution Detection
- Diffusion Model Based Posterior Sampling for Noisy Linear Inverse Problems
- Fast Sampling of Diffusion Models via Operator Learning
- Not Just Pretty Pictures: Text-to-Image Generators Enable Interpretable Interventions for Robust Representations
- Consistent Diffusion Models: Mitigating Sampling Drift by Learning to be Consistent
- Nezha: Deployable and High-Performance Consensus Using Synchronized Clocks
- On Calibrating Diffusion Probabilistic Models
- From paintbrush to pixel: A review of deep neural networks in AI-generated art
- Importance of Aligning Training Strategy with Evaluation for Diffusion Models in 3D Multiclass Segmentation
- Inversion by Direct Iteration: An Alternative to Denoising Diffusion for Image Restoration
- CompoDiff: Versatile Composed Image Retrieval With Latent Diffusion
- CrowdDiff: Multi-Hypothesis Crowd Density Estimation Using Diffusion Models
- ∞-Diff: Infinite Resolution Diffusion with Subsampled Mollified States
- A Comprehensive Survey on Knowledge Distillation of Diffusion Models
- NaturalSpeech 2: Latent Diffusion Models are Natural and Zero-Shot Speech and Singing Synthesizers
- TESS: Text-to-Text Self-Conditioned Simplex Diffusion
Related papers
- TC-VAE: Uncovering Out-of-Distribution Data Generative Factors
- Generative Model for Person Re-Identification: A Review
- Towards Understanding the Interplay of Generative Artificial Intelligence and the Internet
- Are generative approaches to ZSAR a look in the right direction?
- A Comprehensive Review of the Latest Advancements in Large Generative AI Models
- Generating Realistic Blood-Cell Images using Cycle-Consistent Generative Adversial Networks
- Dual-Teacher Class-Incremental Learning With Data-Free Generative Replay
- Deep Dexterous Grasping of Novel Objects from a Single View