1-2 of 2 Results

  • Keywords: deep neural networks x
Clear all

Article

Deep Neural Networks in Computational Neuroscience  

Tim C. Kietzmann, Patrick McClure, and Nikolaus Kriegeskorte

The goal of computational neuroscience is to find mechanistic explanations of how the nervous system processes information to give rise to cognitive function and behavior. At the heart of the field are its models, that is, mathematical and computational descriptions of the system being studied, which map sensory stimuli to neural responses and/or neural to behavioral responses. These models range from simple to complex. Recently, deep neural networks (DNNs) have come to dominate several domains of artificial intelligence (AI). As the term “neural network” suggests, these models are inspired by biological brains. However, current DNNs neglect many details of biological neural networks. These simplifications contribute to their computational efficiency, enabling them to perform complex feats of intelligence, ranging from perceptual (e.g., visual object and auditory speech recognition) to cognitive tasks (e.g., machine translation), and on to motor control (e.g., playing computer games or controlling a robot arm). In addition to their ability to model complex intelligent behaviors, DNNs excel at predicting neural responses to novel sensory stimuli with accuracies well beyond any other currently available model type. DNNs can have millions of parameters, which are required to capture the domain knowledge needed for successful task performance. Contrary to the intuition that this renders them into impenetrable black boxes, the computational properties of the network units are the result of four directly manipulable elements: input statistics, network structure, functional objective, and learning algorithm. With full access to the activity and connectivity of all units, advanced visualization techniques, and analytic tools to map network representations to neural data, DNNs represent a powerful framework for building task-performing models and will drive substantial insights in computational neuroscience.

Article

Visual Perception in the Human Brain: How the Brain Perceives and Understands Real-World Scenes  

Clemens G. Bartnik and Iris I. A. Groen

How humans perceive and understand real-world scenes is a long-standing question in neuroscience, cognitive psychology, and artificial intelligence. Initially, it was thought that scenes are constructed and represented by their component objects. An alternative view proposed that scene perception starts by extracting global features (e.g., spatial layout) first and individual objects in later stages. A third framework focuses on how the brain not only represents objects and layout but how this information combines to allow determining possibilities for (inter)action that the environment offers us. The discovery of scene-selective regions in the human visual system sparked interest in how scenes are represented in the brain. Experiments using functional magnetic resonance imaging show that multiple types of information are encoded in the scene-selective regions, while electroencephalography and magnetoencephalography measurements demonstrate links between the rapid extraction of different scene features and scene perception behavior. Computational models such as deep neural networks offer further insight by how training networks on different scene recognition tasks results in the computation of diagnostic features that can then be tested for their ability to predict activity in human brains when perceiving a scene. Collectively, these findings suggest that the brain flexibly and rapidly extracts a variety of information from scenes using a distributed network of brain regions.