Personal Homepage Calibration Audit

Audit date: 2026-09-06

This audit precedes the next implementation phase. It evaluates the deployed Spatial Intelligence redesign as a personal academic homepage, not as a code or visual regression audit.

Current visitor impression

The page is accurate and polished, but the first impression is still closer to a research statement / job-talk opening than to a personal academic homepage. The hero makes the research positioning visually dominant, while the person is reduced to an eyebrow, a portrait, and a long title. The unified scene is a strong explanation of Spatial Intelligence, but its size and the subsequent trajectory section make the site feel like a research program microsite.

Current structure and measured proportion

Current DOM order:

Hero -> About -> Research Vision -> Featured Research -> Research Trajectory
      -> Publications -> Highlights

Local production browser measurements with reduced motion enabled:

The page contains approximately 8,913 characters of main content. Research Vision, Featured Research, and Research Trajectory together account for about 4,438px at 1440px, compared with 1,463px for Hero + About. This ratio is the primary calibration problem.

Identity findings

• Positive: the hero already includes Yimeng Liu, Ph.D. Candidate, Michigan State University, portrait, CV, and a concise Spatial Intelligence positioning.
• Needs improvement: the H1 is a three-line research-program statement at desktop and consumes most of the mobile first screen; the person’s name is only an eyebrow.
• Needs improvement: email is plain footer text, not a direct mailto: entry; Scholar, GitHub, and CV are mostly icon or lower-page actions rather than a clear identity rail in the hero.
• Needs improvement: About is structurally second, but its personal context is not visible until after the oversized research hero.

Research findings

• The Spatial Intelligence definition is accurate, but the page currently presents the full five-part framework, six-step scene, four evidence cards, and a separate past/present/future timeline. Together they over-explain the same research arc.
• The dynamic scene should remain as one concise visual proof, but it should occupy less vertical attention and be introduced after a short personal introduction.
• The distinction between published foundations, current research, and future agenda is already honest and should be preserved while shortened.

Required target structure

Person
  -> short personal introduction / identity rail
Research
  -> concise Spatial Intelligence positioning and one visual
Selected Work
  -> Hydra, Proteus, Adonis, Driving as evidence of accumulated capability
Publications
  -> complete data-driven scholarly record
Updates
  -> personal academic milestones and current activity
About / Activities
  -> advisor, education, service, talks, collaboration, contact

About remains available as a direct section and in the global navigation, but the first content after the hero should feel like meeting Yimeng before entering the research explanation. A compact About / identity block may remain second to preserve the existing personal-homepage decision.

Phase boundary

This audit makes no source changes to the page. The next implementation phase should first resize and rewrite the identity surface, move meaningful updates earlier, and compress the Research Vision / trajectory copy. It should not introduce another visual system or another research framework.

Audit acceptance

• The current page is measurable as research-heavy rather than merely described as such.
• The personal identity gaps are explicit: hero hierarchy, direct contact links, and introduction placement.
• The existing research truth and Spatial Intelligence mainline are preserved as the research layer, not removed.