Seminars

Within the last decade, Deep Artificial Neural Networks (DNNs) have emerged as powerful computer vision systems that match or exceed human performance on many benchmark tasks such as image classification. But whether current DNNs are suitable computational models of the human visual system remains an open question: While DNNs have proven to be capable of predicting neural activations in primate visual cortex, psychophysical experiments have shown behavioral differences between DNNs and human subjects, as quantified by error consistency. Error consistency is typically measured by briefly presenting natural or corrupted images to human subjects and asking them to perform an n-way classification task under time pressure. But for how long should stimuli ideally be presented to guarantee a fair comparison with DNNs? Here we investigate the influence of presentation time on error consistency, to test the hypothesis that higher-level processing drives behavioral differences. We systematically vary presentation times of backward-masked stimuli from 8.3ms to 266ms and measure human performance and reaction times on natural, lowpass-filtered and noisy images. Our experiment constitutes a fine-grained analysis of human image classification under both image corruptions and time pressure, showing that even drastically time-constrained humans who are exposed to the stimuli for only two frames, i.e. 16.6ms, can still solve our 8-way classification task with success rates way above chance. We also find that human-to-human error consistency is already stable at 16.6ms.

Using videos from the Deepfake Detection Challenge (cf. Groh et al., 2021), we investigated human deepfake detection performance (DDP) in two unique observer groups: Super-Recognizers (SRs) and "normal" officers from within the 18K members of the Berlin Police. SRs were identified either via previously proposed lab-based procedures (Ramon, 2021) or the only existing tool for SR identification involving increasingly challenging, authentic forensic material: beSure® (Berlin Test For Super-Recognizer Identification; Ramon & Rjosk, 2022). Across two experiments we examined deepfake detection performance (DDP) in participants who judged single videos and pairs of videos in a 2AFC decision setting. We explored speed-accuracy trade-offs in DDP, compared DDP between lab-identified SRs and non-SRs, and police officers whose face identity processing skills had been extensively tested using challenging. In this talk I will discuss our surprising findings and argue that further work is needed too determine whether face identity processing is related to DDP or not.

Trends in NeuroAI is a reading group hosted by the MedARC Neuroimaging & AI lab (https://medarc.ai/fmri). This will be an informal journal club presentation, we do not have an author of the paper joining us. Title: Brain decoding: toward real-time reconstruction of visual perception Abstract: In the past five years, the use of generative and foundational AI systems has greatly improved the decoding of brain activity. Visual perception, in particular, can now be decoded from functional Magnetic Resonance Imaging (fMRI) with remarkable fidelity. This neuroimaging technique, however, suffers from a limited temporal resolution (≈0.5 Hz) and thus fundamentally constrains its real-time usage. Here, we propose an alternative approach based on magnetoencephalography (MEG), a neuroimaging device capable of measuring brain activity with high temporal resolution (≈5,000 Hz). For this, we develop an MEG decoding model trained with both contrastive and regression objectives and consisting of three modules: i) pretrained embeddings obtained from the image, ii) an MEG module trained end-to-end and iii) a pretrained image generator. Our results are threefold: Firstly, our MEG decoder shows a 7X improvement of image-retrieval over classic linear decoders. Second, late brain responses to images are best decoded with DINOv2, a recent foundational image model. Third, image retrievals and generations both suggest that MEG signals primarily contain high-level visual features, whereas the same approach applied to 7T fMRI also recovers low-level features. Overall, these results provide an important step towards the decoding - in real time - of the visual processes continuously unfolding within the human brain. Speaker: Dr. Paul Scotti (Stability AI, MedARC) Paper link: https://arxiv.org/abs/2310.19812

Algonauts 2023 was a challenge to create the best model that predicts fMRI brain activity given a seen image. Huze team dominated the competition and released a preprint detailing their process. This journal club meeting will involve open discussion of the paper with Q/A with Huze. Paper: https://arxiv.org/pdf/2308.01175.pdf Related paper also from Huze that we can discuss: https://arxiv.org/pdf/2307.14021.pdf

The mapping from the visual field to V1 can be approximated by a log-polar transform. In this domain, scale is a left-right shift, and rotation is an up-down shift. When fed into a standard shift-invariant convolutional network, this provides scale and rotation invariance. However, translation invariance is lost. In our model, this is compensated for by multiple fixations on an object. Due to the high concentration of cones in the fovea with the dropoff of resolution in the periphery, fully 10 degrees of visual angle take up about half of V1, with the remaining 170 degrees (or so) taking up the other half. This layout provides the basis for the central and peripheral pathways. Simulations with this model closely match human performance in scene classification, and competition between the pathways leads to the peripheral pathway being used for this task. Remarkably, in spite of the property of rotation invariance, this model can explain the inverted face effect. We suggest that the standard method of using image coordinates is the wrong prior for models of primate vision.

Humans are very good at visually recognizing materials and inferring their properties. Without touching surfaces, we can usually tell what they would feel like, and we enjoy vivid visual intuitions about how they typically behave. This is impressive because the retinal image that the visual system receives as input is the result of complex interactions between many physical processes. Somehow the brain has to disentangle these different factors. I will present some recent work in which we show that an unsupervised neural network trained on images of surfaces spontaneously learns to disentangle reflectance, lighting and shape. However, the disentanglement is not perfect, and we find that as a result the network not only predicts the broad successes of human gloss perception, but also the specific pattern of errors that humans exhibit on an image-by-image basis. I will argue this has important implications for thinking about appearance and vision more broadly.

Deep learning techniques have revolutionized the field of image analysis and played a disruptive role in the ability to quickly and efficiently train image analysis models that perform as well as human beings. This talk will cover the beginnings of the application of deep learning in the field of ophthalmology and vision science, and cover a variety of applications of using deep learning as a method for scientific discovery and latent associations.

I describe a well-tested Python module that does automated alignment and warping of faces images, and some advantages over existing solutions. An additional tool I’ve developed does automated extraction of facial features, which can be used in a number of interesting ways. I illustrate the value of wavelet-based features with a brief description of 2 recent studies: perceptual in-painting, and the robustness of the whole-part advantage across a large stimulus set. Finally, I discuss the suitability of various deep learning models for generating stimuli to study perceptual face spaces. I believe those interested in the forensic aspects of face perception may find this talk useful.

Progress in deep learning has spawned great successes in many engineering applications. As a prime example, convolutional neural networks, a type of feedforward neural networks, are now approaching – and sometimes even surpassing – human accuracy on a variety of visual recognition tasks. In this talk, however, I will show that these neural networks and their recent extensions exhibit a limited ability to solve seemingly simple visual reasoning problems involving incremental grouping, similarity, and spatial relation judgments. Our group has developed a recurrent network model of classical and extra-classical receptive field circuits that is constrained by the anatomy and physiology of the visual cortex. The model was shown to account for diverse visual illusions providing computational evidence for a novel canonical circuit that is shared across visual modalities. I will show that this computational neuroscience model can be turned into a modern end-to-end trainable deep recurrent network architecture that addresses some of the shortcomings exhibited by state-of-the-art feedforward networks for solving complex visual reasoning tasks. This suggests that neuroscience may contribute powerful new ideas and approaches to computer science and artificial intelligence.

Object recognition relies on the complex visual representations in cortical areas at the top of the ventral stream hierarchy. While these are thought to be derived from low-level stages of visual processing, this has not been shown, yet. Here, I describe the results of two projects exploring the contributions of primary visual cortex (V1) processing to object recognition using artificial neural networks (ANNs). First, we developed hundreds of ANN-based V1 models and evaluated how their single neurons approximate those in the macaque V1. We found that, for some models, single neurons in intermediate layers are similar to their biological counterparts, and that the distributions of their response properties approximately match those in V1. Furthermore, we observed that models that better matched macaque V1 were also more aligned with human behavior, suggesting that object recognition is derived from low-level. Motivated by these results, we then studied how an ANN’s robustness to image perturbations relates to its ability to predict V1 responses. Despite their high performance in object recognition tasks, ANNs can be fooled by imperceptibly small, explicitly crafted perturbations. We observed that ANNs that better predicted V1 neuronal activity were also more robust to adversarial attacks. Inspired by this, we developed VOneNets, a new class of hybrid ANN vision models. Each VOneNet contains a fixed neural network front-end that simulates primate V1 followed by a neural network back-end adapted from current computer vision models. After training, VOneNets were substantially more robust, outperforming state-of-the-art methods on a set of perturbations. While current neural network architectures are arguably brain-inspired, these results demonstrate that more precisely mimicking just one stage of the primate visual system leads to new gains in computer vision applications and results in better models of the primate ventral stream and object recognition behavior.

One challenge in exploring the internal representation of faces is the lack of controlled stimuli transformations. Researchers are often limited to verbalizable transformations in the creation of a dataset. An alternative approach to verbalization for interpretability is finding image-based measures that allow us to quantify image transformations. In this study, we explore whether PCA could be used to create controlled transformations to a face by testing the effect of these transformations on human perceptual similarity and on computational differences in Gabor, Pixel and DNN spaces. We found that perceptual similarity and the three image-based spaces are linearly related, almost perfectly in the case of the DNN, with a correlation of 0.94. This provides a controlled way to alter the appearance of a face. In experiment 2, the effect of familiarity on the perception of multidimensional transformations was explored. Our findings show that there is a positive relationship between the number of components transformed and both the perceptual similarity and the same three image-based spaces used in experiment 1. Furthermore, we found that familiar faces are rated more similar overall than unfamiliar faces. That is, a change to a familiar face is perceived as making less difference than the exact same change to an unfamiliar face. The ability to quantify, and thus control, these transformations is a powerful tool in exploring the factors that mediate a change in perceived identity.

Not all individuals are equally competent at recognizing the faces they interact with. Revealing how the brains of different individuals support variations in this ability is a crucial step to develop an understanding of real-world human visual behaviour. In this talk, I will present findings from a large high-density EEG dataset (>100k trials of participants processing various stimulus categories) and computational approaches which aimed to characterise the brain representations behind real-world proficiency of “super-recognizers”—individuals at the top of face recognition ability spectrum. Using decoding analysis of time-resolved EEG patterns, we predicted with high precision the trial-by-trial activity of super-recognizers participants, and showed that evidence for face recognition ability variations is disseminated along early, intermediate and late brain processing steps. Computational modeling of the underlying brain activity uncovered two representational signatures supporting higher face recognition ability—i) mid-level visual & ii) semantic computations. Both components were dissociable in brain processing-time (the first around the N170, the last around the P600) and levels of computations (the first emerging from mid-level layers of visual Convolutional Neural Networks, the last from a semantic model characterising sentence descriptions of images). I will conclude by presenting ongoing analyses from a well-known case of acquired prosopagnosia (PS) using similar computational modeling of high-density EEG activity.

State-of-the-art machine vision models can predict human recognition memory for complex scenes with astonishing accuracy. In this talk I present work that investigated how memorable scenes are actually remembered and experienced by human observers. We found that memorable scenes were recognized largely based on recollection of specific episodic details but also based on familiarity for an entire scene. I thus highlight current limitations in machine vision models emulating human recognition memory, with promising opportunities for future research. Moreover, we were interested in what observers specifically remember about complex scenes. We thus considered the functional role of eye-movements as a window into the content of memories, particularly when observers recollected specific information about a scene. We found that when observers formed a memory representation that they later recollected (compared to scenes that only felt familiar), the overall extent of exploration was broader, with a specific subset of fixations clustered around later to-be-recollected scene content, irrespective of the memorability of a scene. I discuss the critical role that our viewing behavior plays in visual memory formation and retrieval and point to potential implications for machine vision models predicting the content of human memories.

Classically, vision is seen as a cascade of local, feedforward computations. This framework has been tremendously successful, inspiring a wide range of ground-breaking findings in neuroscience and computer vision. Recently, feedforward Convolutional Neural Networks (ffCNNs), inspired by this classic framework, have revolutionized computer vision and been adopted as tools in neuroscience. However, despite these successes, there is much more to vision. I will present our work using visual crowding and related psychophysical effects as probes into visual processes that go beyond the classic framework. In crowding, perception of a target deteriorates in clutter. We focus on global aspects of crowding, in which perception of a small target is strongly modulated by the global configuration of elements across the visual field. We show that models based on the classic framework, including ffCNNs, cannot explain these effects for principled reasons and identify recurrent grouping and segmentation as a key missing ingredient. Then, we show that capsule networks, a recent kind of deep learning architecture combining the power of ffCNNs with recurrent grouping and segmentation, naturally explain these effects. We provide psychophysical evidence that humans indeed use a similar recurrent grouping and segmentation strategy in global crowding effects. In crowding, visual elements interfere across space. To study how elements interfere over time, we use the Sequential Metacontrast psychophysical paradigm, in which perception of visual elements depends on elements presented hundreds of milliseconds later. We psychophysically characterize the temporal structure of this interference and propose a simple computational model. Our results support the idea that perception is a discrete process. Together, the results presented here provide stepping-stones towards a fuller understanding of the visual system by suggesting architectural changes needed for more human-like neural computations.