Search

A deep neural network trained only on visual object recognition develops tuned number detectors reminiscent of real neurons., Humans and animals have a “number sense,” an innate capability to intuitively assess the number of visual items in a set, its numerosity. This capability implies that mechanisms to extract numerosity indwell the brain’s visual system, which is primarily concerned with visual object recognition. Here, we show that network units tuned to abstract numerosity, and therefore reminiscent of real number neurons, spontaneously emerge in a biologically inspired deep neural network that was merely trained on visual object recognition. These numerosity-tuned units underlay the network’s number discrimination performance that showed all the characteristics of human and animal number discriminations as predicted by the Weber-Fechner law. These findings explain the spontaneous emergence of the number sense based on mechanisms inherent to the visual system.

In primates, the magnitude system resides in a fronto-parietal network. Single neurons in the monkey prefrontal cortex (PFC) and ventral intraparietal area (VIP) exhibit higher responses to a certain number of stimulus items regardless of their appearance or even sensory modality. Neuroimaging studies in humans show corresponding activation in human fronto-parietal areas during enumeration tasks. However, these areas are also involved in many other executive functions and, thus, the responses of single neurons within the network could be shaped by many factors. Understanding how information about magnitude develops within single neurons in this network was the objective of this thesis. This thesis includes five studies addressing various aspects of the primate frontoparietal magnitude system. First, we determined the role of behavioural relevance in shaping neuronal responses to number. Using enumerable coloured stimuli that naïve macaque monkeys discriminated based on their colour rather than number, we examined the selectivity of neuronal responses towards the number of stimuli. We simultaneously recorded single neurons in VIP and PFC. We compared these neurons to those recorded after a period of training for both monkeys, while they discriminated the stimuli based on number. In all the recording sessions, we also mapped the visual receptive fields (RF) of neurons using a passive fixation task. We created RF maps for a large number of spatially-selective neurons in each area and compared the RFs of pairs of neurons recorded at the same electrode tip. We then differentiated the extent of interaction between the RF and number selectivity in both areas. Neurons in both PFC and VIP were selective for number despite the monkeys being numerically-naïve and number being the behaviourally irrelevant stimulus feature. Post training, neurons in PFC were modulated by behavioural relevance and their selectivity for number became stronger as a result. VIP neurons did not show such an effect. We found that PFC RFs were predominantly contralateral and VIP RFs, foveal. Regardless of RF location and size, we observed heterogeneous and sometimes, inverted RFs in neurons adjacent to each other, more frequently in PFC than in VIP. Lastly, neurons in both PFC and VIP were strongly number-selective even when the number stimuli were shown outside their RFs. Our results provided valuable insight into the organisation of the magnitude system in primates. The presence of number-selective neuronal responses in numerically-naïve monkeys even when the number of stimuli was behaviourally irrelevant confirmed that our magnitude system processes magnitude spontaneously as a natural category. The strict spatiotopic organisation of RFs characteristic of early visual areas is progressively lower in VIP and PFC. Together, these results point to a hierarchy in the fronto-parietal areas we studied, with PFC located at the apex of the magnitude system and VIP upstream to it.

Cognitive impairments prevent older adults from using powered wheelchairs because of safety concerns, thus reducing mobility and resulting in increased dependence on caregivers. An intelligent powered wheelchair system (NOAH) is proposed to help restore mobility, while ensuring safety. Machine vision and learning techniques are described to help prevent collisions with obstacles, and provide reminders and navigation assistance through adaptive audio prompts. The intelligent wheelchair is initially tested in various controlled environments and simulated scenarios. Finally, the system is tested with older adults with mild-to-moderate cognitive impairment through a single-subject research design. Results demonstrate the high diversity of the target population, and highlight the need for customizable assistive technologies that account for the varying capabilities and requirements of the intended users. We show that the collision avoidance module is able to improve safety for all users by lowering the number of frontal collisions. In addition, the wayfinding module assists users in navigating along shorter routes to the destination. Prompting accuracy is found to be quite high during the study. While compliance with correct prompts is high across all users, we notice a distinct difference in the rates of compliance with incorrect prompts. Results show that users who are unsure about the optimal route rely more highly on system prompts for assistance, and thus are able to improve their wayfinding performance by following correct prompts. Improvements in wheelchair position estimation accuracy and joystick usability will help improve user performance and satisfaction. Further user studies will help refine user needs and hopefully allow us to increase mobility and independence of several elderly residents.