Figure Ground Reversal Ambiguity Detector

Audit actual independently observed figure-ground alternatives, not infer perception from arbitrary pixels. Independently annotated feature columns in order: two figure-ground interpretations independently witnessed, same contours independently established. Compatible classes are not established identities; unknown evidence retained. No automatic semantic parsing, diagnosis, authenticity or safety determination.

Description

A figure-ground reversal reading depends on independently witnessed alternatives using the same contours.

A figure-ground reversal reading depends on independently witnessed alternatives using the same contours.^[1]

This tool checks supplied feature annotations without analyzing an image or mind.

When to use Figure Ground Reversal Ambiguity Detector

Use this tool to check supplied feature annotations without analyzing an image or mind.

How to use Figure Ground Reversal Ambiguity Detector and how it works

Independently assess each listed feature or criterion in the fixed order; record unknown evidence with a zero known flag and a valid binary observation placeholder.
Enter observedFeatureFlags, featureKnownFlags and run the tool.
Read result.compatibleClasses, result.witnessedClasses, result.unknownFeatureIndices using the meanings and sign conventions below.
Review missing and unknown evidence separately; resolve it through independent observation rather than treating compatible classes or placeholders as established facts.

Input field	Meaning	Value type	Requirement
observedFeatureFlags	Binary 0-or-1 feature observations in the fixed column order below.	array of number	Required; branch conditions below
featureKnownFlags	Binary 0-or-1 known flags; unknown placeholders do not establish or refute a class.	array of number	Required; branch conditions below

Enter observed-feature and evidence-known binary arrays in the exact feature order below. Require the fixed feature count and integer 0-or-1 flags, including valid placeholders for unknown observations. Each rule uses 1 for a required present feature, 0 for a required absent feature and -1 for an unconstrained feature. A class is compatible when every known constrained feature matches; it is witnessed only when every constrained feature is known and matches. Unknown placeholders do not decide compatibility. Return compatible and witnessed class names in fixed rule order plus zero-based unknown indices. Classes may overlap, and an empty witnessed list is a successful result rather than an input error.

compatible(rule)=all(rule_j==-1 or known_j==0 or observed_j==rule_j); witnessed(rule)=all(rule_j==-1 or (known_j==1 and observed_j==rule_j))

Use the stated ordering, units and branch conditions.

Zero-based column	Meaning
0	two figure-ground interpretations independently witnessed
1	same contours independently established

Class token	Fixed rule in column order
reversal-ambiguity-candidate	[1, 1]

Output field	Meaning	Value type
result	Object containing the fields below.	object
result.compatibleClasses	Fixed rule names not contradicted by known constrained feature evidence.	array of string
result.witnessedClasses	Fixed rule names whose constrained features are all known and satisfied.	array of string
result.unknownFeatureIndices	Zero-based indices whose evidence-known flag is zero.	array

json

{
  "input": {
    "observedFeatureFlags": [
      1,
      1
    ],
    "featureKnownFlags": [
      1,
      1
    ]
  },
  "output": {
    "result": {
      "compatibleClasses": [
        "reversal-ambiguity-candidate"
      ],
      "witnessedClasses": [
        "reversal-ambiguity-candidate"
      ],
      "unknownFeatureIndices": []
    }
  }
}

Limitations and assumptions

The evidence annotations are supplied by the caller. Unknown observations remain unknown, and compatible readings can overlap. The tool does not analyze images or texts, collect observations, prove a theory, certify a design or determine artistic quality.

Supply explicit values and dense aligned arrays. Unexpected fields, wrong types and nonfinite scalar values are rejected; fixed-token, alignment and other branch conditions are explained above. A successful false flag, empty match list or unknown-evidence list differs from an input error. Extremely large inputs or nonrepresentable results may fail.

Alternative or Complementary methods

Use Gestalt Closure Law Completion Checker for a related operation: Incomplete contours and reported completion are separate pieces of evidence about a viewing experience.

Use Gestalt Laws Similarity Proximity Continuity Matrix for a related operation: Several Gestalt grouping principles can be annotated for the same pair of elements.

References

Figure Ground Reversal Ambiguity Detector — tool.io

Similar or alternative tools

Gestalt Closure Law Completion Checker
Incomplete contours and reported completion are separate pieces of evidence about a viewing experience.
Gestalt Laws Similarity Proximity Continuity Matrix
Several Gestalt grouping principles can be annotated for the same pair of elements.