Tracking multiple independent targets: evidence for a parallel tracking mechanism

Area of Science:

• Cognitive Psychology
• Neuroscience
• Visual Perception

Background:

• Visual attention is often considered to be focused at a single point in the visual field.
• This attentional locus can be shifted independently of eye movements.
• Previous models suggest serial processing for certain attentional tasks.

Purpose of the Study:

• To investigate whether any visual attention operations can occur in parallel across multiple locations.
• To test the hypothesis that indexing and tracking object identity can be a parallel process.
• To evaluate the FINST (Fingers of Instantiation) model of parallel object tracking.

Main Methods:

• Two experiments were conducted involving visual search tasks with multiple identical, randomly moving objects.
• Participants were tasked with tracking a subset of objects and identifying changes within the subset.
• Performance was measured by accuracy in distinguishing target changes from distractor changes under varying display parameters.

Main Results:

• Subjects successfully tracked a subset of up to 5 objects among 10, distinguishing target from distractor changes.
• High accuracy (87%) was achieved even when serial scanning models predicted significantly lower performance (approx. 40%).
• These results indicate a parallel processing capability for indexing and maintaining object identities.

Conclusions:

• Certain aspects of visual attention, specifically feature indexing and identity tracking, can operate in parallel across multiple object loci.
• The findings support the FINST model's proposal of a primitive mechanism for parallel maintenance of object indexes.
• This parallel indexing mechanism is crucial for binding visual features, facilitating subsequent pattern recognition.