Joyca Lacroix

Modeling Recognition Memory Using the Similarity Structure of Natural Input

The natural input memory (NIM) model is a new model for recognition memory that operates on natural visual input. A biologically informed perceptual preprocessing method takes local samples (eye fixations) from a natural image and translates these into a feature-vector representation. During recognition, the model compares incoming preprocessed natural input to stored representations. By complementing the recognition memory process with a perceptual front end, the NIM model is able to make predictions about memorability based directly on individual natural stimuli. We demonstrate that the NIM model is able to simulate experimentally obtained similarity ratings and recognition memory for individual stimuli (i.e., face images).

Toward a visual cognitive system using active top-down saccadic control

The saccadic selection of relevant visual input for preferential processing allows the efficient use of computational resources. Based on saccadic active human vision, we aim to develop a plausible saccade-based visual cognitive system for a humanoid robot. This paper presents two initial steps toward our objective by extending the saccade-based model of human memory called NIM1 to a plausible model of natural visual classification. NIM builds feature-vector representations from selected local image samples and uses these to make memory-based decisions. As a first step, we adapt NIM to a straightforward saccade-based model for the classification of natural visual input called NIM-CLASS and evaluate the model in a face-classification experiment. As a second step, we aim to approach the interactive nature of human vision by extending NIM-CLASS to NIM-CLASSTD by adding active top-down saccadic control. We then assess to what extent top-down control enhances classification performance. The results show that the incorporation of top-down saccadic control benefits classification performance compared to the purely bottom-up control, reducing the amount of visual input required for correct classification. We conclude that NIM-CLASSTD may provide a fruitful basis for an active visual cognitive system in a humanoid robot that enables efficient visual processing.

The Development of Small-world Semantic Networks

Cognitive processes rely on knowledge structures that can be represented by networks of interconnected concepts, i.e., semantic networks. We study the developmental dynamics of generic semantic-network models. We focus on two measures: (i) the characteristic path length between an arbitrary pair of concepts, and (ii) the clustering coefficient of groups of concepts. Short path lengths facilitate the dynamics of mental processes through spreading activation and a large clustering coefficient reflects the existence of structured representations. We analysed semantic-network models generated from behavioural data. We measure the characteristic path length and clustering coefficient at various stages of development. The results reveal semantic networks to be small-world networks, i.e., networks that combine short path lengths with high clustering. In addition, developing semantic networks are characterised by an increase in small-worldliness by maintaining a constant path length and clustering coefficient despite the increase in the number of concepts.