mlsanityData to 7 Oct 2026

Latent Representation Matters: Human-like Sketches in One-shot Drawing Tasks

arXiv 2406.06079 ↗First posted 10 Jun 2024cs.CVcited by 3

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

Abstract (arXiv)

Humans can effortlessly draw new categories from a single exemplar, a feat that has long posed a challenge for generative models. However, this gap has started to close with recent advances in diffusion models. This one-shot drawing task requires powerful inductive biases that have not been systematically investigated. Here, we study how different inductive biases shape the latent space of Latent Diffusion Models (LDMs). Along with standard LDM regularizers (KL and vector quantization), we explore supervised regularizations (including classification and prototype-based representation) and contrastive inductive biases (using SimCLR and redundancy reduction objectives). We demonstrate that LDMs with redundancy reduction and prototype-based regularizations produce near-human-like drawings (regarding both samples' recognizability and originality) -- better mimicking human perception (as evaluated psychophysically). Overall, our results suggest that the gap between humans and machines in one-shot drawings is almost closed.

Key facts. Cited by 3 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 36 of its 72 references that are in the map. First posted on arXiv 10 Jun 2024.

Builds on

36 of 72 references.

Closest ancestors · 1

Same direction, newest first

  1. 2023Diffusion Models as Artists: Are we Closing the Gap between Humans and Machines?27 Jan

Other references · 35

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  3. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  4. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
  5. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,482
  6. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,005
  7. 2022Classifier-Free Diffusion Guidancecited by 5,717
  8. 2022Generative Adversarial Networkscited by 5,603
  9. 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,712
  10. 2017Neural Discrete Representation Learningcited by 4,650
  11. 2019Generative Modeling by Estimating Gradients of the Data Distributioncited by 4,067
  12. 2017Prototypical Networks for Few-shot Learningcited by 3,001
23 more
  1. 2016Matching Networks for One Shot Learningcited by 2,246
  2. 2021Barlow Twins: Self-Supervised Learning via Redundancy Reductioncited by 1,818
  3. 2017Learning to Compare: Relation Network for Few-Shot Learningcited by 1,139
  4. 2020Prototypical Contrastive Learning of Unsupervised Representationscited by 636
  5. 2021With a Little Help from My Friends: Nearest-Neighbor Contrastive Learning of Visual Representationscited by 324
  6. 2017Data Augmentation Generative Adversarial Networkscited by 278
  7. 2016Variational Lossy Autoencodercited by 175
  8. 2019From Variational to Deterministic Autoencoderscited by 131
  9. 2016Towards a Neural Statisticiancited by 119
  10. 2022SinDDM: A Single Image Denoising Diffusion Modelcited by 60
  11. 2019The Omniglot challenge: a 3-year progress reportcited by 54
  12. 2016One-Shot Generalization in Deep Generative Modelscited by 44
  13. 2022SinDiffusion: Learning a Diffusion Model from a Single Natural Imagecited by 42
  14. 2022Few-Shot Diffusion Modelscited by 37
  15. 2023General Image-to-Image Translation with One-Shot Image Guidancecited by 31
  16. 2018The Variational Homoencoder: Learning to learn high capacity generative models from few examplescited by 23
  17. 2021Deep Metric Learning for Few-Shot Image Classification: A Review of Recent Developmentscited by 23
  18. 2023End-To-End Latent Variational Diffusion Models for Inverse Problems in High Energy Physicscited by 22
  19. 2021Predify: Augmenting deep neural networks with brain-inspired predictive coding dynamicscited by 19
  20. 2023SEVA: Leveraging sketches to evaluate alignment between human and machine visual abstractioncited by 14
  21. 2022Diversity vs. Recognizability: Human-like generalization in one-shot generative modelscited by 9
  22. 2017Semi-Supervised and Active Few-Shot Learning with Prototypical Networkscited by 8
  23. 2020Iterative VAE as a predictive brain model for out-of-distribution generalizationcited by 6

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 2; 2026: 1.

3 papers.

  1. 2025Follow the Energy, Find the Path: Riemannian Metrics from Energy-Based Modelscited by 7
  2. 2025Stable diffusion models reveal a persisting human and AI gap in visual creativitycited by 0
  3. 2026Not Too Generative, Not Too Discriminative: The Human Alignment Sweet Spotcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024An analytic theory of creativity in convolutional diffusion models
  2. 2024It's All About Your Sketch: Democratising Sketch Control in Diffusion Models
  3. 2026Diffusing in the Right Space: A Systematic Study of Latent Diffusability
  4. 2025The Evolving Nature of Latent Spaces: From GANs to Diffusion
  5. 2024KnobGen: Controlling the Sophistication of Artwork in Sketch-Based Diffusion Models
  6. 2026VideoSketcher: Sequential Sketch Generation Using Video Model Priors
  7. 2023Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
  8. 2025Color encoding in Latent Space of Stable Diffusion Models