Guy Gaziv, Sarah Goulding, Ani Ayvazian-Hancock, Yoon Bai, James DiCarlo

arXiv

Precise control of neural activity -- modulating target neurons deep in the brain while leaving nearby neurons unaffected -- is an outstanding challenge in neuroscience, generally approached using invasive techniques. This study investigates the possibility of precisely and noninvasively modulating neural activity in the high-level primate ventral visual stream via perturbations on one's natural visual feed. When tested on macaque inferior temporal (IT) neural populations, we found quantitative agreement between the model-predicted and biologically realized effect: strong modulation concentrated on targeted neural sites. We extended this to demonstrate accurate injection of experimenter-chosen neural population patterns via subtle perturbations applied on the background of typical natural visual feeds. These results highlight that current machine-executable models of the ventral stream can now design noninvasive, visually-delivered, possibly imperceptible neural interventions at the resolution of individual neurons.

Morgan Talbot, Gabriel Kreiman, James DiCarlo, Guy Gaziv

International Conference on Learning Representations (ICLR) 2025

The currently leading artificial neural network (ANN) models of the visual ventral stream -- which are derived from a combination of performance optimization and robustification methods -– have demonstrated a remarkable degree of behavioral alignment with humans on visual categorization tasks. Extending upon previous work, we show that not only can these models guide image perturbations that change the induced human category percepts, but they also can enhance human ability to accurately report the original ground truth. Furthermore, we find that the same models can also be used out-of-the-box to predict the proportion of correct human responses to individual images, providing a simple, human-aligned estimator of the relative difficulty of each image. Motivated by these observations, we propose to augment visual learning in humans in a way that improves human categorization accuracy at test time. Our learning augmentation approach consists of (i) selecting images based on their model-estimated recognition difficulty, and (ii) using image perturbations that aid recognition for novice learners. We find that combining these model-based strategies gives rise to test-time categorization accuracy gains of 33-72% relative to control subjects without these interventions, despite using the same number of training feedback trials. Surprisingly, beyond the accuracy gain, the training time for the augmented learning group was also shorter by 20-23%. We demonstrate the efficacy of our approach in a fine-grained categorization task with natural images, as well as tasks in two clinically relevant image domains -- histology and dermoscopy -- where visual learning is notoriously challenging. To the best of our knowledge, this is the first application of ANNs to successfully increase visual learning performance in humans, and especially robustly across varied image domains.

Guy Gaziv*, Michael Lee*, James DiCarlo

Neural Information Processing Systems (NeurIPS) 2023

The visual object category reports of artificial neural networks (ANNs) are notoriously sensitive to tiny, adversarial image perturbations. Because human category reports (aka human percepts) are thought to be insensitive to those same small-norm perturbations – and locally stable in general – this argues that ANNs are incomplete scientific models of human visual perception. Consistent with this, we show that when small-norm image perturbations are generated by standard ANN models, human object category percepts are indeed highly stable. However, in this very same "human-presumed-stable" regime, we find that robustified ANNs reliably discover low-norm image perturbations that strongly disrupt human percepts. These previously undetectable human perceptual disruptions are massive in amplitude, approaching the same level of sensitivity seen in robustified ANNs. Further, we show that robustified ANNs support precise perceptual state interventions: they guide the construction of low-norm image perturbations that strongly alter human category percepts toward specific prescribed percepts. In sum, these contemporary models of biological visual processing are now accurate enough to guide strong and precise interventions on human perception.

Ganit Kupershmidt*, Roman Beliy*, Guy Gaziv, Michal Irani

arXiv

Reconstructing natural videos from fMRI brain recordings is very challenging, for two main reasons: (i) As fMRI data acquisition is diffcult, we only have a limited amount of supervised samples, which is not enough to cover the huge space of natural videos; and (ii) The temporal resolution of fMRI recordings is much lower than the frame rate of natural videos. In this paper, we propose a selfsupervised approach for natural movie reconstruction. By employing cycle consistency over Encoding-Decoding natural videos, we can: (i) exploit the full framerate of the training videos, and not be limited only to clips that correspond to fMRI recordings; (ii) exploit massive amounts of external natural videos which the subjects never saw inside the fMRI machine. These enable increasing the applicable training data by several orders of magnitude, introducing natural video priors to the decoding network, as well as temporal coherence. Our approach signifcantly outperforms competing methods, since those train only on the limited supervised data. We further introduce a new and simple temporal prior of natural videos, which when folded into our fMRI decoder further allows us to reconstruct videos at a higher framerate (HFR) of up to x8 of the original fMRI sample rate.

Guy Gaziv, Michal Irani

arXiv

In the past few years, significant advancements were made in reconstruction of observed natural images from fMRI brain recordings using deep-learning tools. Here, for the first time, we show that dense 3D depth maps of observed 2D natural images can also be recovered directly from fMRI brain recordings. We use an off-the-shelf method to estimate the unknown depth maps of natural images. The estimated depth maps are then used as an auxiliary reconstruction criterion to train for depth reconstruction directly from fMRI. We propose two main approaches: Depth-only recovery and joint image-depth RGBD recovery. Because the number of available "paired" training data (images with fMRI) is small, we enrich the training data via self-supervised cycle-consistent training on many "unpaired" data (natural images & depth maps without fMRI). This is achieved using our newly defined and trained Depth-based Perceptual Similarity metric as a reconstruction criterion. We show that predicting the depth map directly from fMRI outperforms its indirect sequential recovery from the reconstructed images. We further show that activations from early cortical visual areas dominate our depth reconstruction results, and propose means to characterize fMRI voxels by their degree of depth-information tuning. This work adds an important layer of decoded information, extending the current envelope of visual brain decoding capabilities.

Guy Gaziv*, Roman Beliy*, Niv Granot*, Assaf Hoogi, Francesca Strappini, Tal Golan, Michal Irani

NeuroImage 2022

Reconstructing natural images and decoding their semantic category from fMRI brain recordings is challenging. Acquiring sufficient pairs of images and their corresponding fMRI responses, which span the huge space of natural images, is prohibitive. We present a novel self-supervised approach that goes well beyond the scarce paired data, for achieving both: (i) state-of-the art fMRI-to-image reconstruction, and (ii) first-ever large-scale semantic classification from fMRI responses. By imposing cycle consistency between a pair of deep neural networks (from image-to-fMRI & from fMRI-to-image), we train our image reconstruction network on a large number of "unpaired" natural images (images without fMRI recordings) from many novel semantic categories. This enables to adapt our reconstruction network to a very rich semantic coverage without requiring any explicit semantic supervision. Specifically, we find that combining our self-supervised training with high-level perceptual losses, gives rise to new reconstruction & classification capabilities. In particular, this perceptual training enables to classify well fMRIs of never-before-seen semantic classes, without requiring any class labels during training. This gives rise to: (i) Unprecedented image-reconstruction from fMRI of never-before-seen images (evaluated by image metrics and human testing), and (ii) Large-scale semantic classification of categories that were never-before-seen during network training. Such large-scale (1000-way) semantic classification from fMRI recordings has never been demonstrated before. Finally, we provide evidence for the biological consistency of our learned model.

Shany Grossman*, Guy Gaziv*, Erin M Yeagle, Michal Harel, Pierre Mégevand, David M Groppe, Simon Khuvis, Jose L Herrero, Michal Irani, Ashesh D Mehta, Rafael Malach

Nature Communications 2019

The discovery that deep convolutional neural networks (DCNNs) achieve human performance in realistic tasks offers fresh opportunities for linking neuronal tuning properties to such tasks. Here we show that the face-space geometry, revealed through pair-wise activation similarities of face-selective neuronal groups recorded intracranially in 33 patients, significantly matches that of a DCNN having human-level face recognition capabilities. This convergent evolution of pattern similarities across biological and artificial networks highlights the significance of face-space geometry in face perception. Furthermore, the nature of the neuronal to DCNN match suggests a role of human face areas in pictorial aspects of face perception. First, the match was confined to intermediate DCNN layers. Second, presenting identity-preserving image manipulations to the DCNN abolished its correlation to neuronal responses. Finally, DCNN units matching human neuronal group tuning displayed view-point selective receptive fields. Our results demonstrate the importance of face-space geometry in the pictorial aspects of human face perception.

Roman Beliy*, Guy Gaziv*, Assaf Hoogi, Francesca Strappini, Tal Golan, Michal Irani

Neural Information Processing Systems (NeurIPS) 2019

Reconstructing observed images from fMRI brain recordings is challenging. Unfortunately, acquiring sufficient ''labeled'' pairs of {Image, fMRI} (i.e., images with their corresponding fMRI responses) to span the huge space of natural images is prohibitive for many reasons. We present a novel approach which, in addition to the scarce labeled data (training pairs), allows to train fMRI-to-image reconstruction networks also on "unlabeled" data (i.e., images without fMRI recording, and fMRI recording without images). The proposed model utilizes both an Encoder network (image-to-fMRI) and a Decoder network (fMRI-to-image). Concatenating these two networks back-to-back (Encoder-Decoder & Decoder-Encoder) allows augmenting the training data with both types of unlabeled data. Importantly, it allows training on the unlabeled test-fMRI data. This self-supervision adapts the reconstruction network to the new input test-data, despite its deviation from the statistics of the scarce training data.

Guy Gaziv, Lior Noy, Yuvalal Liron, Uri Alon

PLOS One 2017

Face-to-face conversations are central to human communication and a fascinating example of joint action. Beyond verbal content, one of the primary ways in which information is conveyed in conversations is body language. Body motion in natural conversations has been difficult to study precisely due to the large number of coordinates at play. There is need for fresh approaches to analyze and understand the data, in order to ask whether dyads show basic building blocks of coupled motion. Here we present a method for analyzing body motion during joint action using depth-sensing cameras, and use it to analyze a sample of scientific conversations. Our method consists of three steps: defining modes of body motion of individual participants, defining dyadic modes made of combinations of these individual modes, and lastly defining motion motifs as dyadic modes that occur significantly more often than expected given the single-person motion statistics. As a proof-of-concept, we analyze the motion of 12 dyads of scientists measured using two Microsoft Kinect cameras. In our sample, we find that out of many possible modes, only two were motion motifs: synchronized parallel torso motion in which the participants swayed from side to side in sync, and still segments where neither person moved. We find evidence of dyad individuality in the use of motion modes. For a randomly selected subset of 5 dyads, this individuality was maintained for at least 6 months. The present approach to simplify complex motion data and to define motion motifs may be used to understand other joint tasks and interactions. The analysis tools developed here and the motion dataset are publicly available.