Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

Images with harder-to-reconstruct visual representations leave stronger memory traces

Nature Human Behaviour (2024)Cite this article

Subjects

Abstract

Much of what we remember is not because of intentional selection, but simply a by-product of perceiving. This raises a foundational question about the architecture of the mind: how does perception interface with and influence memory? Here, inspired by a classic proposal relating perceptual processing to memory durability, the level-of-processing theory, we present a sparse coding model for compressing feature embeddings of images, and show that the reconstruction residuals from this model predict how well images are encoded into memory. In an open memorability dataset of scene images, we show that reconstruction error not only explains memory accuracy, but also response latencies during retrieval, subsuming, in the latter case, all of the variance explained by powerful vision-only models. We also confirm a prediction of this account with ‘model-driven psychophysics’. This work establishes reconstruction error as an important signal interfacing perception and memory, possibly through adaptive modulation of perceptual processing.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Model architecture and example images.
Fig. 2: Images with a large reconstruction error are more memorable (Nimages = 2,221).
Fig. 3: Images with a large reconstruction error are recognized faster during retrieval (Nimages = 2,221).
Fig. 4: Example images from each of the four groups with different distinctiveness–reconstruction error profiles.
Fig. 5: Images with harder-to-reconstruct representations benefit more from longer encoding times (Nparticipants = 45).

Similar content being viewed by others

Data availability

Data used in Studies 1 and 2 are from a publicly available dataset from Isola et al.12 (https://web.mit.edu/phillipi/Public/MemorabilityPAMI/index.html). De-identified data collected for Study 3 have been deposited on GitHub (https://github.com/CNCLgithub/ReconMem)68.

Code availability

Codes have been deposited on GitHub (https://github.com/CNCLgithub/ReconMem)68.

References

  1. Wagner, A. D. et al. Building memories: remembering and forgetting of verbal experiences as predicted by brain activity. Science 281, 1188–1191 (1998).

    Article  CAS  PubMed  Google Scholar 

  2. Xue, G. The neural representations underlying human episodic memory. Trends Cogn. Sci. 22, 544–561 (2018).

    Article  PubMed  Google Scholar 

  3. Craik, F. I. & Lockhart, R. S. Levels of processing: a framework for memory research. J. Verbal Learning Verbal Behav. 11, 671–684 (1972).

    Article  Google Scholar 

  4. Schurgin, M. W., Wixted, J. T. & Brady, T. F. Psychophysical scaling reveals a unified theory of visual memory strength. Nat. Hum. Behav. 4, 1156–1172 (2020).

    Article  PubMed  Google Scholar 

  5. Chun, M. M. & Johnson, M. K. Memory: enduring traces of perceptual and reflective attention. Neuron 72, 520–535 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Kurby, C. A. & Zacks, J. M. Segmentation in the perception and memory of events. Trends Cogn. Sci. 12, 72–79 (2008).

    Article  PubMed  PubMed Central  Google Scholar 

  7. Favila, S. E., Lee, H. & Kuhl, B. A. Transforming the concept of memory reactivation. Trends Neurosci. 43, 939–950 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Liu, J. et al. Transformative neural representations support long-term episodic memory. Sci. Adv. 7, eabg9715 (2021).

    Article  PubMed  PubMed Central  Google Scholar 

  9. Libby, A. & Buschman, T. J. Rotational dynamics reduce interference between sensory and memory representations. Nat. Neurosci. 24, 715–726 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Serences, J. T. Neural mechanisms of information storage in visual short-term memory. Vision Res. 128, 53–67 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  11. Xu, Y. Reevaluating the sensory account of visual working memory storage. Trends Cogn. Sci. 21, 794–815 (2017).

    Article  PubMed  Google Scholar 

  12. Isola, P., Xiao, J., Parikh, D., Torralba, A. & Oliva, A. What makes a photograph memorable? IEEE Trans. Pattern Anal. Mach. Intell. 7, 1469–1482 (2014).

    Article  Google Scholar 

  13. Bainbridge, W. A., Isola, P. & Oliva, A. The intrinsic memorability of face photographs. J. Exp. Psychol. Gen. 142, 1323–1334 (2013).

    Article  PubMed  Google Scholar 

  14. Jaegle, A. et al. Population response magnitude variation in inferotemporal cortex predicts image memorability. eLife 8, e47596 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Khosla, A., Raju, A. S., Torralba, A. & Oliva, A. Understanding and predicting image memorability at a large scale. In Proc. IEEE International Conference on Computer Vision, 2390–2398 (2015).

  16. Lin, Q., Yousif, S. R., Scholl, B. & Chun, M. M. Image memorability is driven by visual and conceptual distinctivenes. J. Vis. 19, 290c (2019).

    Article  Google Scholar 

  17. Kramer, M. A., Hebart, M. N., Baker, C. I. & Bainbridge, W. A. The features underlying the memorability of objects. Sci. Adv. 9, eadd2981 (2023).

    Article  PubMed  PubMed Central  Google Scholar 

  18. Baddeley, A. D. The trouble with levels: a reexamination of Craik and Lockhart’s framework for memory research. Psychol. Rev. 85, 139–152 (1978).

    Article  Google Scholar 

  19. Treisman, A. in Levels of Processing in Human Memory (eds Cermak, L. S. & Craik, F. I. M.) 301–330 (Psychology Press, 2014).

  20. Craik, F. I. Remembering: an activity of mind and brain. Annu. Rev. Psychol. 71, 1–24 (2020).

    Article  PubMed  Google Scholar 

  21. Cermak, L. S. & Craik, F. I. M. Levels of Processing in Human Memory (Psychology Press, 2014).

  22. Bainbridge, W. A. The resiliency of image memorability: a predictor of memory separate from attention and priming. Neuropsychologia 141, 107408 (2020).

    Article  PubMed  Google Scholar 

  23. Bates, C. J. & Jacobs, R. A. Efficient data compression in perception and perceptual memory. Psychol. Rev. 127, 891–917 (2020).

    Article  PubMed  Google Scholar 

  24. Schacter, D. L. Adaptive constructive processes and the future of memory. Am. Psychol. 67, 603–613 (2012).

    Article  PubMed  Google Scholar 

  25. Hemmer, P. & Steyvers, M. A Bayesian account of reconstructive memory. Top. Cogn. Sci. 1, 189–202 (2009).

    Article  PubMed  Google Scholar 

  26. Olshausen, B. A. & Field, D. J. Emergence of simple-cell receptive field properties by learning a sparse code for natural images. Nature 381, 607–609 (1996).

    Article  CAS  PubMed  Google Scholar 

  27. Olshausen, B. A. & Field, D. J. Sparse coding with an overcomplete basis set: a strategy employed by v1? Vision Res. 37, 3311–3325 (1997).

    Article  CAS  PubMed  Google Scholar 

  28. Benna, M. K. & Fusi, S. Place cells may simply be memory cells: memory compression leads to spatial tuning and history dependence. Proc. Natl Acad. Sci. USA 118, e2018422118 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Lewicki, M. S. Efficient coding of natural sounds. Nat. Neurosci. 5, 356–363 (2002).

    Article  CAS  PubMed  Google Scholar 

  30. Zemel, R. & Hinton, G. E. Developing population codes by minimizing description length. Adv. Neural Info. Process. Syst. 6, 11–18 (1993).

    Google Scholar 

  31. Rozell, C. J., Johnson, D. H., Baraniuk, R. G. & Olshausen, B. A. Sparse coding via thresholding and local competition in neural circuits. Neural Comput. 20, 2526–2563 (2008).

    Article  PubMed  Google Scholar 

  32. Rumelhart, D., Hinton, G. & Williams, R. Learning representations by back-propagating errors. Nature 323, 533–536 (1986).

    Article  Google Scholar 

  33. Gregor, K. & LeCun, Y. Learning fast approximations of sparse coding. In Proc. 27th International Conference on International Conference on Machine Learning, 399–406 (2010).

  34. Zhou, B., Lapedriza, A., Khosla, A., Oliva, A. & Torralba, A. Places: a 10 million image database for scene recognition. IEEE Trans. Pattern Anal. Mach. Intell. 40, 1452–1464 (2017).

    Article  PubMed  Google Scholar 

  35. Simonyan, K. & Zisserman, A. Very deep convolutional networks for large-scale image recognition. In Proc. of the 3rd International Conference on Learning Representations 1–14 (ICLR, 2015).

  36. Berger, T. Rate Distortion Theory: A Mathematical Basis for Data Compression (Prentice-Hall, 1971).

  37. Cover, T. M. & Thomas, J. A. Elements of Information Theory (Wiley, 1991).

  38. MacKay, D. J. Information Theory, Inference, and Learning Algorithms (Cambridge Univ. Press, 2003).

  39. Hinton, G. E., Dayan, P., Frey, B. J. & Neal, R. M. The ‘wake–sleep’ algorithm for unsupervised neural networks. Science 268, 1158–1161 (1995).

    Article  CAS  PubMed  Google Scholar 

  40. Kahana, M. & Loftus, G. in The Nature of Cognition (ed. Sternberg, R. J.) 322–384 (MIT Press, 1999).

  41. Bylinskii, Z., Isola, P., Bainbridge, C., Torralba, A. & Oliva, A. Intrinsic and extrinsic effects on image memorability. Vision Res. 116, 165–178 (2015).

    Article  PubMed  Google Scholar 

  42. Vincent, A., Craik, F. I. & Furedy, J. J. Relations among memory performance, mental workload and cardiovascular responses. Int. J. Psychophysiol. 23, 181–198 (1996).

    Article  CAS  Google Scholar 

  43. Ragland, J. D. et al. Levels-of-processing effect on word recognition in schizophrenia. Biol. Psychiatry 54, 1154–1161 (2003).

    Article  PubMed  PubMed Central  Google Scholar 

  44. Broers, N., Potter, M. C. & Nieuwenstein, M. R. Enhanced recognition of memorable pictures in ultra-fast RSVP. Psychon. Bull. Rev. 25, 1080–1086 (2018).

    Article  PubMed  Google Scholar 

  45. Craik, F. I. Levels of processing: past, present… and future? Memory 10, 305–318 (2002).

    Article  PubMed  Google Scholar 

  46. Friston, K. & Kiebel, S. Predictive coding under the free-energy principle. Philos. Trans. R. Soc. B 364, 1211–1221 (2009).

    Article  Google Scholar 

  47. Rosenbaum, R. On the relationship between predictive coding and backpropagation. PLoS ONE 17, e0266102 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Barrow, H. G. & Tenenbaum, J. M. In Computer Vision Systems (eds. Hanson A. & Riseman E. M.) 3–26 (Academic Press, 1978).

  49. Olshausen, B. A., Mangun, G. & Gazzaniga, M. Perception as an Inference Problem (MIT Press, 2014).

  50. Yuille, A. & Kersten, D. Vision as Bayesian inference: analysis by synthesis? Trends Cogn. Sci. 10, 301–308 (2006).

    Article  PubMed  Google Scholar 

  51. Mumford, D. in Large-Scale Neuronal Theories of the Brain (eds Koch, C & Davis, J. L.) 125–152 (MIT Press, 1994).

  52. Brewer, J. B., Zhao, Z., Desmond, J. E., Glover, G. H. & Gabrieli, J. D. Making memories: brain activity that predicts how well visual experience will be remembered. Science 281, 1185–1187 (1998).

    Article  CAS  PubMed  Google Scholar 

  53. Paller, K. A. & Wagner, A. D. Observing the transformation of experience into memory. Trends Cogn. Sci. 6, 93–102 (2002).

    Article  PubMed  Google Scholar 

  54. Kim, H. Neural activity that predicts subsequent memory and forgetting: a meta-analysis of 74 fMRI studies. Neuroimage 54, 2446–2461 (2011).

    Article  PubMed  Google Scholar 

  55. Xue, G. et al. Greater neural pattern similarity across repetitions is associated with better memory. Science 330, 97–101 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  56. Ward, E. J., Chun, M. M. & Kuhl, B. A. Repetition suppression and multi-voxel pattern similarity differentially track implicit and explicit visual memory. J. Neurosci. 33, 14749–14757 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  57. Voss, J. L., Bridge, D. J., Cohen, N. J. & Walker, J. A. A closer look at the hippocampus and memory. Trends Cogn. Sci. 21, 577–588 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  58. Ryan, J. D., Shen, K. & Liu, Z.-X. The intersection between the oculomotor and hippocampal memory systems: empirical developments and clinical implications. Ann. N Y Acad. Sci. 1464, 115–141 (2020).

    Article  PubMed  Google Scholar 

  59. Kragel, J. E. & Voss, J. L. Looking for the neural basis of memory. Trends Cogn. Sci. 26, 53–65 (2022).

    Article  PubMed  Google Scholar 

  60. Lyu, M. et al. Overt attentional correlates of memorability of scene images and their relationships to scene semantics. J. Vis. 20, 1–17 (2020).

    Article  Google Scholar 

  61. Cohendet, R., Demarty, C.-H., Duong, N. Q. & Engilberge, M. Videomem: constructing, analyzing, predicting short-term and long-term video memorability. In Proc. IEEE/CVF International Conference on Computer Vision, 2531–2540 (2019).

  62. Xu, Q., Fang, F., Molino, A., Subbaraju, V. & Lim, J.-H. Predicting event memorability from contextual visual semantics. Adv. Neural Info. Process. Syst. 34, 22431–22442 (2021).

    Google Scholar 

  63. Lau, M. C., Goh, W. D. & Yap, M. J. An item-level analysis of lexical-semantic effects in free recall and recognition memory using the megastudy approach. Q. J. Exp. Psychol. (Hove) 71, 2207–2222 (2018).

    Article  PubMed  Google Scholar 

  64. Majumdar, A. et al. Where are we in the search for an artificial visual cortex for embodied intelligence? Adv. Neural Info. Process. Syst. 36, 1–23 (2024).

    Google Scholar 

  65. Radford, A. et al. Language models are unsupervised multitask learners. OpenAI Blog 1, 9 (2019).

    Google Scholar 

  66. Stahl, A. E. & Feigenson, L. Observing the unexpected enhances infants’ learning and exploration. Science 348, 91–94 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  67. Chollet, F. et al. Keras. https://keras.io (2015).

  68. Lin, Q., Li, Z., Lafferty, J. & Yildirim, I. From seeing to remembering: Images with harder-to-reconstruct representations leave stronger memory traces. GitHub https://github.com/CNCLgithub/ReconMem (2023).

Download references

Acknowledgements

This project was funded by an Air Force Office of Scientific Research (AFOSR) award #FA9550-22-1-0041 (to I.Y.). The funder had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript. We thank the Yale Center for Research Computing for maintaining HPC resources for computation. We also thank R. Jacobs, B. Scholl and members of the Yale Cognitive & Neural Computation Lab for comments on an earlier version of this manuscript.

Author information

Author notes

  1. These authors contributed equally: Qi Lin, Zifan Li.

  2. These authors jointly supervised this work: John Lafferty, Ilker Yildirim.

Authors and Affiliations

  1. Department of Psychology, Yale University, New Haven, CT, USA

    Qi Lin & Ilker Yildirim

  2. Center for Brain Science, RIKEN, Wako, Japan

    Qi Lin

  3. Department of Statistics & Data Science, Yale University, New Haven, CT, USA

    Zifan Li, John Lafferty & Ilker Yildirim

  4. Wu-Tsai Institute, Yale University, New Haven, CT, USA

    John Lafferty & Ilker Yildirim

Authors
  1. Qi Lin
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Zifan Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. John Lafferty
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Ilker Yildirim
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Q.L., J.L. and I.Y. conceived the study. Q.L., Z.L., J.L. and I.Y. developed the methodology. Q.L. and Z.L. developed the software. Q.L. collected the data. Q.L. and Z.L. formally analysed the data. Q.L., Z.L. and I.Y. wrote the original draft. Q.L., Z.L., J.L. and I.Y. wrote and edited the manuscript. Q.L. visualized the data. J.L. and I.Y. supervised the study. I.Y. acquired the funding.

Corresponding authors

Correspondence to Qi Lin, John Lafferty or Ilker Yildirim.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Human Behaviour thanks the anonymous reviewers for their contribution to the peer review of this work. Peer reviewer reports are available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Lin, Q., Li, Z., Lafferty, J. et al. Images with harder-to-reconstruct visual representations leave stronger memory traces. Nat Hum Behav (2024). https://doi.org/10.1038/s41562-024-01870-3

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1038/s41562-024-01870-3

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing