Funding

Some funding may be available here. The School of Psychology offers funding for a limited number of PhD studentships. Details on how to apply for a PhD studentship can be found here. If you wish to apply for funding from the Scool of Psychology please note interest by emailing me mwo@st-andrews.ac.uk so we can discuss possible projects etc.

Overview of projects

The projects suitable for PhD students are focussed around my research interests. While I, like many other neuroscientists, are interested in the "big question" - how does the brain produce the mind, such a question is far too big to hope for anything like a complete answer. But we can and are making inroads into this most fundamental of problems. Some possible PhD project areas are listed below.

If you are interested in persuing any of these or other PhD projects with me, please email mwo@st-andrews.ac.uk.

Perception of actions: The analysis of complex movements, and in particular of body movements and actions, is a central function of the visual system that can be crucial for survival. Movement recognition is essential for the analysis of the behavior of other living agents, and gestures and facial movements provide important information channels for social communication. At the same time, action recognition forms a critical step of the learning of motor behavior by imitation..

While several relevant cortical areas have been identified, an understanding of the detailed underlying processes is largely lacking. Fundamental questions remain unanswered: how are the ventral and the dorsal pathways interacting to integrate form and motion features? Is the recognition of actions dependent on predictive representations that anticipate future action states or goals over time, and what are the circuits at the level of individual neurons that implement such prediction? Finally, how is the information about body movements and shape and position of goal objects integrated?

Potential starting points include

1. Identification of critical “snap-shot” images: It will be tested how the interpretation of ambiguous static images (e.g., picking up or putting down of an object) is influenced by effector-object and effector-body relationships. By introduction of local motion information (by presentation of frame pairs) we will determine how motion interacts with such static images. Results will be compared with predictions from hierarchical neural models for the recognition of “snap-shot” images and motion patterns.
2. Identification of cue hierarchies: Preliminary data indicates the existence of a hierarchy of static cues in action perception (e.g., hand position/orientation, gaze direction, head orientation). We will study the interaction between such static cues for different types of actions, resulting in constraints for feature hierarchies in computational models.

Just-in-time processing: Understanding the visual system will have large scientific
and practical value, yet despite a large research effort this goal is still not reached. In the past our failure to implement a reliable vision system in silicon might have been blamed on lack of computing power, but with increasing computer power it has become clear that our lack of understanding of the visual system is a far greater obstacle.

To make progress we need to understand how the nonlinearities and complexities of the visual system interact to operate successfully over a wide range of circumstances. We have developed a novel model of neural coding and
computation, just-in-time processing, in which the system uses neural non-linearities to work across varying stimulus contrasts and conditions. Just-in-time processing allows circuits involved in vision to change seamlessly between a
fast feedforwardmode under ideal conditions, and a slow feedback dominated mode under poor signal conditions (van Rossum et al, 2008).

This model accurately predicts response timing as a function of contrast, and also explains several empirical phenomena previously thought to be unrelated. We suggest that the just-in-time principle applies to many aspects of vision and possibly sensory processing in general, and will also have practical applications.
This proposal combines theoretical models, visual psychophysics, and detailed analysis of single unit recordings done in the primate. By bringing these disciplines to mutual benefit and to better understand vision as an adaptive system.

Computational studies will be performed under the co-supervision of Mark van Rossum (ANC, Edinburgh).

Memories and their interactions. Given the importance attached to memory in everyday life, the inability to recall items on demand can be problematic. An apparently ironic phenomenon has been identified however which suggests that in addition to retrieving desired memories, the act of remembering inhibits or suppresses related memories (Anderson & McCulloch 1999; Anderson & Spellman 1995; Anderson et al. 1994, 2001; Bjorket al. 1998; Ciranni & Shimamura 1999; MacLeod & Macrae 2001; Macrae & MacLeod 1999).

A competitive model, designed to investigate the development of the cortical visual system (Foldiak 1990, 1991; Oram & foldiak 1996), provides an explanation for the suppression of some memories as a consequence of remembering others (Oram, in prep). Some specific predictions based on this model as to when retrieval-induced forgetting effects should or should not occur have already been examined.

The model suggests that the mechanisms by which memories are formed and adapted may also underlie retrieval-induced forgetting effects.

Perceptual anticipation. Electrophysiological recordings investigating action encoding in the STS (Oram & Perrett 1994, 1996; Barraclough et al 2005, 2006) and the processing of natural image sequences (Perrett et al in press), combined with a diffusion model (Oram et al 2002; Oram 2005;) result in predictions that can be tested (Oram & Dritschel, in prep; Wattie & Oram, in prep).

We have recently shown that ordered sequences of images facilitate perception of subsequent predictable images, but disrupt the perception of unpredictable images (Perrett et al, in press). To continue this line of work participants can be asked to report on the presence or absence of a particular image whilst manipulating the immediately preceding visual experience. Potential starting points include

1. Ordered versus random sequences will be tested, varying the visual as well as the conceptual similarity between image and history (e.g. by changing the identity of an actor or action goals).
2. In addition, the influence of the speed and smoothness of the motion can be investigated.