NeurIPS 2023 accepted papers on Image generation

253 of the 3,584 papers accepted at NeurIPS 2023 (7.1%) are on Image generation, compared with 3.0% at NeurIPS 2022. They fall into 7 topics.

See NeurIPS 2023 in the interactive map

Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.

Topics and papers

Fast diffusion sampling · 87 papers

ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with Variational Score Distillation (159 citations)
Emergent Correspondence from Image Diffusion (53 citations)
A Tale of Two Features: Stable Diffusion Complements DINO for Zero-Shot Semantic Correspondence (49 citations)
DIFUSCO: Graph-based Diffusion Solvers for Combinatorial Optimization (34 citations)
Diffusion Hyperfeatures: Searching Through Time and Space for Semantic Correspondence (33 citations)
UniPC: A Unified Predictor-Corrector Framework for Fast Sampling of Diffusion Models (33 citations)
DatasetDM: Synthesizing Data with Perception Annotations Using Diffusion Models (32 citations)
Patch Diffusion: Faster and More Data-Efficient Training of Diffusion Models (32 citations)
Generate What You Prefer: Reshaping Sequential Recommendation via Guided Diffusion (31 citations)
DiffTraj: Generating GPS Trajectory with Diffusion Probabilistic Model (26 citations)
Diffusion with Forward Models: Solving Stochastic Inverse Problems Without Direct Supervision (23 citations)
Structural Pruning for Diffusion Models (23 citations)

and 75 more in the interactive map.

Autoregressive and latent generation · 60 papers

High-Fidelity Audio Compression with Improved RVQGAN (101 citations)
Simple and Controllable Music Generation (67 citations)
GenImage: A Million-Scale Benchmark for Detecting AI-Generated Image (55 citations)
Tree-Rings Watermarks: Invisible Fingerprints for Diffusion Images (54 citations)
DreamSim: Learning New Dimensions of Human Visual Similarity using Synthetic Data (46 citations)
Voicebox: Text-Guided Multilingual Universal Speech Generation at Scale (45 citations)
Reconstructing the Mind's Eye: fMRI-to-Image with Contrastive Learning and Diffusion Priors (40 citations)
Seeing is not always believing: Benchmarking Human and Model Perception of AI-Generated Images (31 citations)
CRoSS: Diffusion Model Makes Controllable, Robust and Secure Image Steganography (30 citations)
Any-to-Any Generation via Composable Diffusion (29 citations)
Crystal Structure Prediction by Joint Equivariant Diffusion (26 citations)
A Unified Framework for U-Net Design and Analysis (25 citations)

and 48 more in the interactive map.

Text-to-image diffusion · 52 papers

ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generation (108 citations)
Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image Generation (99 citations)
GenEval: An object-focused framework for evaluating text-to-image alignment (80 citations)
BLIP-Diffusion: Pre-trained Subject Representation for Controllable Text-to-Image Generation and Editing (65 citations)
Uni-ControlNet: All-in-One Control to Text-to-Image Diffusion Models (65 citations)
LayoutGPT: Compositional Visual Planning and Generation with Large Language Models (62 citations)
T2I-CompBench: A Comprehensive Benchmark for Open-world Compositional Text-to-image Generation (47 citations)
Subject-driven Text-to-Image Generation via Apprenticeship Learning (44 citations)
Diffusion Self-Guidance for Controllable Image Generation (42 citations)
Hard Prompts Made Easy: Gradient-Based Discrete Optimization for Prompt Tuning and Discovery (40 citations)
RAPHAEL: Text-to-Image Generation via Large Mixture of Diffusion Paths (40 citations)
Optimizing Prompts for Text-to-Image Generation (38 citations)

and 40 more in the interactive map.

Image restoration · 23 papers

Cold Diffusion: Inverting Arbitrary Image Transforms Without Noise (103 citations)
PromptIR: Prompting for All-in-One Image Restoration (85 citations)
Global Structure-Aware Diffusion Process for Low-light Image Enhancement (78 citations)
Lossy Image Compression with Conditional Diffusion Models (30 citations)
Hierarchical Integration Diffusion Model for Realistic Image Deblurring (29 citations)
A Unified Conditional Framework for Diffusion-based Image Restoration (11 citations)
PromptRestorer: A Prompting Image Restoration Method with Degradation Perception (11 citations)
Deep Fractional Fourier Transform (10 citations)
PUCA: Patch-Unshuffle and Channel Attention for Enhanced Self-Supervised Image Denoising (8 citations)
StyleGAN knows Normal, Depth, Albedo, and More (6 citations)
CHAMMI: A benchmark for channel-adaptive models in microscopy imaging (3 citations)
Deep Optimal Transport: A Practical Algorithm for Photo-realistic Image Restoration (3 citations)

and 11 more in the interactive map.

Editing and style transfer · 17 papers

MagicBrush: A Manually Annotated Dataset for Instruction-Guided Image Editing (61 citations)
Understanding the Latent Space of Diffusion Models through the Lens of Riemannian Geometry (13 citations)
CycleNet: Rethinking Cycle Consistency in Text-Guided Diffusion for Image Manipulation (9 citations)
AUDIT: Audio Editing by Following Instructions with Latent Diffusion Models (8 citations)
PHOTOSWAP: Personalized Subject Swapping in Images (7 citations)
Visual Instruction Inversion: Image Editing via Image Prompting (6 citations)
DiffUTE: Universal Text Editing Diffusion Model (5 citations)
Conditional Score Guidance for Text-Driven Image-to-Image Translation (4 citations)
Optimal Transport-Guided Conditional Score-Based Diffusion Model (4 citations)
Face Reconstruction from Facial Templates by Learning Latent Space of a Generator Network (3 citations)
FaceDNeRF: Semantics-Driven Face Reconstruction, Prompt Editing and Relighting with Diffusion Models (3 citations)
IMPRESS: Evaluating the Resilience of Imperceptible Perturbations Against Unauthorized Data Usage in Diffusion-Based Generative AI (3 citations)

and 5 more in the interactive map.