Neuroscience framework for sensorless gaze-following in HRI. Exploits the brain's convexity prior (hollow-face illusion) to create perceived mutual gaze without sensors, power, or computation. Grounded in STS gaze processing, convexity prior, and predictive processing hierarchy. Use when: human-robot interaction, gaze-following design, low-cost robotics, perceptual illusions in HRI, neuroscience-inspired design, child-robot interaction. Trigger: sensorless gaze, hollow-face illusion, gaze-fol...
Scanned 9/11/2026
Install to Claude Code
npx -y skills add hiyenwong/ai_collection --skill perception-neuroscience-framework-sensorless-gaze --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Perception Neuroscience Framework Sensorless Gaze?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/hiyenwong-perception-neuroscience-framework-sensorless-gaze-ai-collection)More formats (shields.io, HTML) on the badges page.
---
name: perception-neuroscience-framework-sensorless-gaze
description: >
Neuroscience framework for sensorless gaze-following in HRI. Exploits the brain's convexity prior
(hollow-face illusion) to create perceived mutual gaze without sensors, power, or computation.
Grounded in STS gaze processing, convexity prior, and predictive processing hierarchy.
Use when: human-robot interaction, gaze-following design, low-cost robotics, perceptual illusions
in HRI, neuroscience-inspired design, child-robot interaction.
Trigger: sensorless gaze, hollow-face illusion, gaze-following HRI, convexity prior,
low-cost robot design, perception-based interaction, 无传感器凝视, 空心脸错觉.
version: 1.0.0
author: Research Synthesis (arXiv:2604.09829)
license: MIT
metadata:
hermes:
tags: [HRI, gaze-following, neuroscience, convexity-illusion, low-cost-robotics, perception]
source_paper: "Perception Is All You Need: A Neuroscience Framework for Low Cost Sensorless Gaze in HRI (arXiv:2604.09829)"
---
# Sensorless Gaze-Following via Neuroscience Framework
## Overview
A sub-dollar cardboard robot design that exploits the human brain's own gaze computation pipeline
in reverse — making the viewer's perceptual system the robot's "actuator" with zero sensors,
zero power, zero computation, and zero privacy concerns.
## Neuroscience Foundation
Three converging mechanisms explain why concave eye sockets with painted pupils produce
perceived mutual gaze from any viewing angle:
### 1. Superior Temporal Sulcus (STS) Gaze Processing
- Distributed face processing network computes gaze direction
- STS is specialized for interpreting eye direction
- Brain assumes eyes are convex (protruding outward)
### 2. High-Precision Convexity Prior
- Brain strongly expects faces to be convex
- Causes the "hollow-face illusion" — concave faces perceived as convex
- When face appears convex, painted pupils appear to track viewer
### 3. Predictive Processing Hierarchy
- Top-down face knowledge overrides bottom-up depth signals
- Prior expectation of convex face > actual concave geometry
- Results in perceived mutual gaze from any angle
## Design Principle
```
Concave Eye Socket + Painted Pupil
↓
Viewer's Brain Applies Convexity Prior
↓
Brain Perceives Convex Face with Forward-Looking Eyes
↓
Perception: "The robot is looking at me"
↓
Mutual gaze achieved — no sensors, no code, no power
```
## Design Constraints (from Perceptual Science)
| Parameter | Constraint | Rationale |
|-----------|------------|-----------|
| Socket depth | Sufficient concavity | Strong hollow-face illusion |
| Pupil placement | Center of concave region | Appears to track from all angles |
| Lighting | Even illumination | Avoids depth cues that break illusion |
| Viewing distance | Typical interaction range | Within illusion effectiveness zone |
## Boundary Conditions
### Will Succeed
- Typical child-adult interaction distances
- Normal lighting conditions
- Neurotypical observers
- Standard viewing angles
### May Fail
- Very young children (developing face processing)
- Clinical populations with face processing differences
- Extreme viewing angles (geometric limits)
- Strong directional lighting (reveals concavity)
## Applications
- Scalable child-robot interaction programs
- Educational robotics at population scale
- Low-cost social robotics research
- Privacy-sensitive environments
- Resource-limited settings
## Implementation Template
```python
# Design parameters (from paper)
GAZE_DESIGN = {
"socket_depth": "concave, ~2-3cm depth",
"pupil_diameter": "1-2cm, centered in socket",
"face_width": "child-appropriate scale (~15cm)",
"material": "cardboard or 3D printed",
"cost": "< $1 USD",
"sensors_required": False,
"power_required": False,
"privacy_risk": "None"
}
```
## Activation Keywords
- sensorless gaze, hollow-face illusion, gaze-following HRI
- convexity prior, low-cost robotics, neuroscience design
- 无传感器凝视, 空心脸错觉, 凝视跟随
## References
- Mason Kadem. "Perception Is All You Need: A Neuroscience Framework for Low Cost Sensorless Gaze
in HRI." arXiv:2604.09829
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!