Cognitive Science Lecture with Judy Fan

Introduction

• Speaker Introduction: Judy Fan, a prominent young cognitive scientist, recognized for creativity and rigor.
• Background: Transitioned from neuroscience to cognitive science, studying perception to complex cognitive processes.
• Research Interests:
  • Artistic and narrative expression
  • Learning and teaching
  • Symbolic understanding and cultural cognition

Cognitive Tools

• Definition: Tools like the number line are human inventions that aid in understanding and processing the world.
• Historical Context:
  • Importance of visual representation in science (e.g., Darwin’s finches, Galileo's telescope).
  • Visual abstraction aids in recognizing patterns and solving problems.

Research Focus

• Goal: Understand how the human mind enables innovation and knowledge representation.
• Frameworks Studied:
  • Cognitive tools and their influence on thought
  • Intersection of science and engineering in cognitive development

Visual Abstraction and Communication

• Visual Perception: Transforming sensory input into meaningful experiences.
• Visual Production: Creating marks that convey meaning.
• Visual Communication:
  • Differentiating between faithful depictions and abstract representations.
  • Studies:
    • Sketch recognition and abstraction using neural networks.
    • Influence of context on the level of detail in drawings.
    • Emergence of new graphical conventions.

Mechanistic Explanations vs. Visual Depictions

• Hypotheses:
  • Cumulative Hypothesis: Explanations are extended depictions.
  • Dissociable Hypothesis: Explanations focus on mechanisms, not appearance.
• Study Results:
  • Explanations prioritize causal mechanisms over visual fidelity.
  • Visual explanations and depictions serve different communicative purposes.

Artificial Systems in Visual Abstraction

• SEVA Benchmark:
  • Evaluating human-like abstraction and understanding in AI.
  • Gaps between human and AI recognition abilities.
• Generative Models:
  • Testing AI models like CLIPasso for sketch generation.
  • Human and AI sketch simplification differs under time constraints.

Data Visualization and Statistical Reasoning

• Importance: Data visualization as a tool for understanding complex phenomena.
• Historical Context: Playfair’s time series as a foundational development.
• Current Challenges:
  • Benchmark Study: Comparing human and AI understanding of data visuals.
  • Model Limitations: Current AI models lag behind humans in interpretation accuracy.

Educational Implications

• Teaching and Learning: Data literacy is crucial for future education.
• Assessment of Skills:
  • Developing better tools and strategies for teaching data visualization.

Conclusion

• Judy Fan’s research aims to bridge cognitive science with practical applications in education and technology.
• Understanding cognitive tools can enhance educational practices and foster innovation.