mlsanityData to 7 Oct 2026

Diffusion in the Dark: A Diffusion Model for Low-Light Text Recognition

arXiv 2303.04291 ↗First posted 7 Mar 2023eess.IV cs.CVcited by 20

On the map: Diffusion models → Image generation, text-to-image and editing

Abstract (arXiv)

Capturing images is a key part of automation for high-level tasks such as scene text recognition. Low-light conditions pose a challenge for high-level perception stacks, which are often optimized on well-lit, artifact-free images. Reconstruction methods for low-light images can produce well-lit counterparts, but typically at the cost of high-frequency details critical for downstream tasks. We propose Diffusion in the Dark (DiD), a diffusion model for low-light image reconstruction for text recognition. DiD provides qualitatively competitive reconstructions with that of state-of-the-art (SOTA), while preserving high-frequency details even in extremely noisy, dark conditions. We demonstrate that DiD, without any task-specific optimization, can outperform SOTA low-light methods in low-light text recognition on real images, bolstering the potential of diffusion models to solve ill-posed inverse problems.

Key facts. Cited by 20 papers in the map (data through 7 Oct 2026). Placed in Image generation, text-to-image and editing, one of 14 directions in Diffusion models. Builds on 73 of its 112 references that are in the map. First posted on arXiv 7 Mar 2023.

Builds on

73 of 112 references.

Closest ancestors · 3

Same direction, newest first

  1. 2022Blended Latent Diffusion6 Jun
  2. 2021GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models20 Dec
  3. 2021Palette: Image-to-Image Diffusion Models10 Nov

Other references · 70

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  3. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  4. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  5. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
  6. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
  7. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,185
  8. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,005
  9. 2016Image-to-Image Translation with Conditional Adversarial Networkscited by 4,510
  10. 2019Generative Modeling by Estimating Gradients of the Data Distributioncited by 4,067
  11. 2022Elucidating the Design Space of Diffusion-Based Generative Modelscited by 2,994
  12. 2016Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Networkcited by 1,898
38 more
  1. 2021Image Super-Resolution via Iterative Refinementcited by 1,506
  2. 2022RePaint: Inpainting using Denoising Diffusion Probabilistic Modelscited by 1,419
  3. 2018Demystifying MMD GANscited by 1,268
  4. 2021Cascaded Diffusion Models for High Fidelity Image Generationcited by 1,155
  5. 2022Diffusion Models in Vision: A Surveycited by 1,008
  6. 2022Denoising Diffusion Restoration Modelscited by 951
  7. 2018Deep Retinex Decomposition for Low-Light Enhancementcited by 720
  8. 2021ILVR: Conditioning Method for Denoising Diffusion Probabilistic Modelscited by 634
  9. 2020Zero-Reference Deep Curve Estimation for Low-Light Image Enhancementcited by 541
  10. 2021Solving Inverse Problems in Medical Imaging with Score-Based Generative Modelscited by 514
  11. 2021Label-Efficient Semantic Segmentation with Diffusion Modelscited by 495
  12. 2019EnlightenGAN: Deep Light Enhancement without Paired Supervisioncited by 466
  13. 2018Learning to See in the Darkcited by 459
  14. 2019Kindling the Darkness: A Practical Low-light Image Enhancercited by 355
  15. 2022Diffusion Models already have a Semantic Latent Spacecited by 324
  16. 2020Retinex-inspired Unrolling with Cooperative Prior Architecture Search for Low-light Image Enhancementcited by 303
  17. 2015A note on the evaluation of generative modelscited by 291
  18. 2021SegDiff: Image Segmentation with Diffusion Probabilistic Modelscited by 279
  19. 2018Unprocessing Images for Learned Raw Denoisingcited by 250
  20. 2022Diffusion Models for Medical Anomaly Detectioncited by 231
  21. 2015LLNet: A Deep Autoencoder Approach to Natural Low-light Image Enhancementcited by 224
  22. 2021Conditional Image Generation with Score-Based Diffusion Modelscited by 195
  23. 2021Low-Light Image Enhancement with Normalizing Flowcited by 184
  24. 2021On Fast Sampling of Diffusion Probabilistic Modelscited by 170
  25. 2017Burst Denoising with Kernel Prediction Networkscited by 156
  26. 2022Scene Text Recognition with Permuted Autoregressive Sequence Modelscited by 148
  27. 2019ICDAR2019 Robust Reading Challenge on Arbitrary-Shaped Text (RRC-ArT)cited by 147
  28. 2020A Physics-based Noise Formation Model for Extreme Low-light Raw Denoisingcited by 138
  29. 2020Towards Accurate Scene Text Recognition with Semantic Reasoning Networkscited by 131
  30. 2021Learning to Efficiently Sample from Diffusion Probabilistic Modelscited by 129
  31. 2019ICDAR 2019 Robust Reading Challenge on Reading Chinese Text on Signboardcited by 127
  32. 2018Dark Model Adaptation: Semantic Image Segmentation from Daytime to Nighttimecited by 122
  33. 2020SEED: Semantics Enhanced Encoder-Decoder Framework for Scene Text Recognitioncited by 119
  34. 2018Scene Text Detection and Recognition: The Deep Learning Eracited by 114
  35. 2017ICDAR2017 Competition on Reading Chinese Text in the Wild (RCTW-17)cited by 103
  36. 2022Accelerating Diffusion Models via Early Stop of the Diffusion Processcited by 100
  37. 2022LEDNet: Joint Low-light Enhancement and Deblurring in the Darkcited by 95
  38. 2020Map-Guided Curriculum Domain Adaptation and Uncertainty-Aware Evaluation for Semantic Nighttime Image Segmentationcited by 86

50 most cited of 70 All 70 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 6; 2024: 5; 2025: 8; 2026: 1.

20 papers.

Sort:

By area:

  1. 2023VideoCrafter1: Open Diffusion Models for High-Quality Video Generationcited by 530
  2. 2024Diffusion Models in Low-Level Vision: A Surveycited by 73
  3. 2024LoLI-Street: Benchmarking Low-Light Image Enhancement and Beyondcited by 9
  4. 2025TSCnet: A Text-driven Semantic-level Controllable Framework for Customized Low-Light Image Enhancementcited by 9
  5. 2025Enhancing Text-to-Image Diffusion Transformer via Split-Text Conditioningcited by 7
  6. 2023DiffSCI: Zero-Shot Snapshot Compressive Imaging via Iterative Spectral Diffusion Modelcited by 6
  7. 2023DiffuseRAW: End-to-End Generative RAW Image Processing for Low-Light Imagescited by 4
  8. 2025Structure-guided Diffusion Transformer for Low-Light Image Enhancementcited by 4
  9. 2025ItDPDM: Information-Theoretic Discrete Poisson Diffusion Modelcited by 4
  10. 2024Zero-Shot Low Light Image Enhancement with Diffusion Priorcited by 3
  11. 2023Tell Me What You See: Text-Guided Real-World Image Denoisingcited by 2
  12. 2023ClassLIE: Structure- and Illumination-Adaptive Classification for Low-Light Image Enhancementcited by 2
8 more
  1. 2025TS-Diff: Two-Stage Diffusion Model for Low-Light RAW Image Enhancementcited by 2
  2. 2025DarkDiff: Advancing Low-Light Raw Enhancement by Retasking Diffusion Models for Camera ISPcited by 2
  3. 2025DRWKV: Focusing on Object Edges for Low-Light Image Enhancementcited by 2
  4. 2024DifuzCam: Replacing Camera Lens with a Mask and a Diffusion Modelcited by 1
  5. 2024Personalized Generative Low-light Image Denoising and Enhancementcited by 1
  6. 2023Dual Degradation-Inspired Deep Unfolding Network for Low-Light Image Enhancementcited by 0
  7. 2025Diffusion Models for Low-Light Image Enhancement: A Multi-Perspective Taxonomy and Performance Analysiscited by 0
  8. 2026Reading in the Dark: Low-light Scene Text Recognitioncited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023DiffusionSTR: Diffusion Model for Scene Text Recognition
  2. 2025Text-Aware Image Restoration with Diffusion Models
  3. 2026LGTM: Training-Free Light-Guided Text-to-Image Diffusion Model via Initial Noise Manipulation
  4. 2023Text-to-Image Diffusion Models are Zero-Shot Classifiers
  5. 2025Dark Noise Diffusion: Noise Synthesis for Low-Light Image Denoising
  6. 2023De-Diffusion Makes Text a Strong Cross-Modal Interface
  7. 2023Diffusion-based Blind Text Image Super-Resolution
  8. 2024InitNO: Boosting Text-to-Image Diffusion Models via Initial Noise Optimization