Reconstructing the Mind's Eye: fMRI-to-Image with Contrastive Learning and Diffusion Priors

• URL: http://arxiv.org/abs/2305.18274v2
• Date: Sat, 7 Oct 2023 12:32:15 GMT
• Title: Reconstructing the Mind's Eye: fMRI-to-Image with Contrastive Learning and Diffusion Priors
• Abstract summary: MindEye is a novel fMRI-to-image approach to retrieve and reconstruct viewed images from brain activity. We show MindEye achieves state-of-the-art performance in both reconstruction and retrieval tasks.
• Score: 5.055354526753589
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We present MindEye, a novel fMRI-to-image approach to retrieve and reconstruct viewed images from brain activity. Our model comprises two parallel submodules that are specialized for retrieval (using contrastive learning) and reconstruction (using a diffusion prior). MindEye can map fMRI brain activity to any high dimensional multimodal latent space, like CLIP image space, enabling image reconstruction using generative models that accept embeddings from this latent space. We comprehensively compare our approach with other existing methods, using both qualitative side-by-side comparisons and quantitative evaluations, and show that MindEye achieves state-of-the-art performance in both reconstruction and retrieval tasks. In particular, MindEye can retrieve the exact original image even among highly similar candidates indicating that its brain embeddings retain fine-grained image-specific information. This allows us to accurately retrieve images even from large-scale databases like LAION-5B. We demonstrate through ablations that MindEye's performance improvements over previous methods result from specialized submodules for retrieval and reconstruction, improved training techniques, and training models with orders of magnitude more parameters. Furthermore, we show that MindEye can better preserve low-level image features in the reconstructions by using img2img, with outputs from a separate autoencoder. All code is available on GitHub.

Related papers

• MIRAGE: Robust multi-modal architectures translate fMRI-to-image models from vision to mental imagery [3.24465146296532]
  MIRAGE is a method explicitly designed to train on vision datasets and cross-decode mental images from brain activity.<n>We show that mental image reconstruction works best when decoders use image features with relatively few dimensions.
  arXiv  Detail & Related papers  (2026-05-16T23:53:43Z)
• Brain-IT: Image Reconstruction from fMRI via Brain-Interaction Transformer [6.890296061697977]
  "Brain-IT" is a brain-inspired approach to image reconstruction from fMRI brain recordings.<n>All model components are shared by all clusters & subjects, allowing efficient training with a limited amount of data.<n>With only 1-hour of fMRI data from a new subject, we achieve results comparable to current methods trained on full 40-hour recordings.
  arXiv  Detail & Related papers  (2025-10-29T21:21:54Z)
• Brain-optimized inference improves reconstructions of fMRI brain activity [0.0]
  We evaluate the prospect of further improving recent decoding methods by optimizing for consistency between reconstructions and brain activity during inference. We sample seed reconstructions from a base decoding method, then iteratively refine these reconstructions using a brain-optimized encoding model. We show that reconstruction quality can be significantly improved by explicitly aligning decoding to brain activity distributions.
  arXiv  Detail & Related papers  (2023-12-12T20:08:59Z)
• Not All Image Regions Matter: Masked Vector Quantization for Autoregressive Image Generation [78.13793505707952]
  Existing autoregressive models follow the two-stage generation paradigm that first learns a codebook in the latent space for image reconstruction and then completes the image generation autoregressively based on the learned codebook. We propose a novel two-stage framework, which consists of Masked Quantization VAE (MQ-VAE) Stack model from modeling redundancy.
  arXiv  Detail & Related papers  (2023-05-23T02:15:53Z)
• Brain Captioning: Decoding human brain activity into images and text [1.5486926490986461]
  We present an innovative method for decoding brain activity into meaningful images and captions. Our approach takes advantage of cutting-edge image captioning models and incorporates a unique image reconstruction pipeline. We evaluate our methods using quantitative metrics for both generated captions and images.
  arXiv  Detail & Related papers  (2023-05-19T09:57:19Z)
• Joint fMRI Decoding and Encoding with Latent Embedding Alignment [77.66508125297754]
  We introduce a unified framework that addresses both fMRI decoding and encoding. Our model concurrently recovers visual stimuli from fMRI signals and predicts brain activity from images within a unified framework.
  arXiv  Detail & Related papers  (2023-03-26T14:14:58Z)
• BrainCLIP: Bridging Brain and Visual-Linguistic Representation Via CLIP for Generic Natural Visual Stimulus Decoding [51.911473457195555]
  BrainCLIP is a task-agnostic fMRI-based brain decoding model. It bridges the modality gap between brain activity, image, and text. BrainCLIP can reconstruct visual stimuli with high semantic fidelity.
  arXiv  Detail & Related papers  (2023-02-25T03:28:54Z)
• Mind Reader: Reconstructing complex images from brain activities [16.78619734818198]
  We focus on reconstructing the complex image stimuli from fMRI (functional magnetic resonance imaging) signals. Unlike previous works that reconstruct images with single objects or simple shapes, our work aims to reconstruct image stimuli rich in semantics. We find that incorporating an additional text modality is beneficial for the reconstruction problem compared to directly translating brain signals to images.
  arXiv  Detail & Related papers  (2022-09-30T06:32:46Z)
• Attentive Symmetric Autoencoder for Brain MRI Segmentation [56.02577247523737]
  We propose a novel Attentive Symmetric Auto-encoder based on Vision Transformer (ViT) for 3D brain MRI segmentation tasks. In the pre-training stage, the proposed auto-encoder pays more attention to reconstruct the informative patches according to the gradient metrics. Experimental results show that our proposed attentive symmetric auto-encoder outperforms the state-of-the-art self-supervised learning methods and medical image segmentation models.
  arXiv  Detail & Related papers  (2022-09-19T09:43:19Z)
• Facial Image Reconstruction from Functional Magnetic Resonance Imaging via GAN Inversion with Improved Attribute Consistency [5.705640492618758]
  We propose a new framework to reconstruct facial images from fMRI data. The proposed framework accomplishes two goals: (1) reconstructing clear facial images from fMRI data and (2) maintaining the consistency of semantic characteristics.
  arXiv  Detail & Related papers  (2022-07-03T11:18:35Z)
• Multiscale Voxel Based Decoding For Enhanced Natural Image Reconstruction From Brain Activity [0.22940141855172028]
  We present a novel approach for enhanced image reconstruction, in which existing methods for object decoding and image reconstruction are merged together. This is achieved by conditioning the reconstructed image to its decoded image category using a class-conditional generative adversarial network and neural style transfer. The results indicate that our approach improves the semantic similarity of the reconstructed images and can be used as a general framework for enhanced image reconstruction.
  arXiv  Detail & Related papers  (2022-05-27T18:09:07Z)
• Is Deep Image Prior in Need of a Good Education? [57.3399060347311]
  Deep image prior was introduced as an effective prior for image reconstruction. Despite its impressive reconstructive properties, the approach is slow when compared to learned or traditional reconstruction techniques. We develop a two-stage learning paradigm to address the computational challenge.
  arXiv  Detail & Related papers  (2021-11-23T15:08:26Z)
• NAS-DIP: Learning Deep Image Prior with Neural Architecture Search [65.79109790446257]
  Recent work has shown that the structure of deep convolutional neural networks can be used as a structured image prior. We propose to search for neural architectures that capture stronger image priors. We search for an improved network by leveraging an existing neural architecture search algorithm.
  arXiv  Detail & Related papers  (2020-08-26T17:59:36Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.