Recognizing the Unseen: A Verified, Multimodal Framework for Trauma-Informed AI
Fuente:
Zenodo
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Recurso digital |
| Lingua: | inglese |
| Pubblicazione: |
Zenodo
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866901667351363584 |
|---|---|
| author | George, Michelle Lynn |
| author_facet | George, Michelle Lynn |
| contents | <p><em>Recognizing the Unseen</em> forms the perceptual and multimodal layer of the Trauma-Informed AI Framework. It brings together visual, auditory, and linguistic information to observe the quiet emotional patterns that often stay hidden—expressions softened by suppression, withdrawal, or guarded restraint. By integrating prosodic features, microexpression dynamics, and transformer-based language representations, the system creates a unified space for reasoning about emotional coherence across multiple forms of data.</p> <p>Through fuzzy-tier calibration (Reflective, Cautious, Assertive) and symbolic verification in Z3, the framework ensures that every interpretation is approached with caution, clarity, and ethical balance. Rather than seeking certainty, it is designed to pause when signals conflict or fade—protecting ambiguity as meaningful evidence rather than error.</p> <p>This work concludes the <em>Trauma-Informed AI Framework Trilogy</em>, following <em>The Haunting Problem</em> (2025) and <em>Empathy as Verification</em> (2025). Together, the three studies establish a connected foundation of theory, logic, and perception for trauma-aware and ethically grounded intelligent systems.</p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_17705532 |
| institution | Zenodo |
| language | eng |
| publishDate | 2025 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Recognizing the Unseen: A Verified, Multimodal Framework for Trauma-Informed AI George, Michelle Lynn trauma-informed AI multimodal emotion recognition affective computing symbolic verification Fuzzy Logic Z3 solver microexpression analysis ethical AI empathy modeling <p><em>Recognizing the Unseen</em> forms the perceptual and multimodal layer of the Trauma-Informed AI Framework. It brings together visual, auditory, and linguistic information to observe the quiet emotional patterns that often stay hidden—expressions softened by suppression, withdrawal, or guarded restraint. By integrating prosodic features, microexpression dynamics, and transformer-based language representations, the system creates a unified space for reasoning about emotional coherence across multiple forms of data.</p> <p>Through fuzzy-tier calibration (Reflective, Cautious, Assertive) and symbolic verification in Z3, the framework ensures that every interpretation is approached with caution, clarity, and ethical balance. Rather than seeking certainty, it is designed to pause when signals conflict or fade—protecting ambiguity as meaningful evidence rather than error.</p> <p>This work concludes the <em>Trauma-Informed AI Framework Trilogy</em>, following <em>The Haunting Problem</em> (2025) and <em>Empathy as Verification</em> (2025). Together, the three studies establish a connected foundation of theory, logic, and perception for trauma-aware and ethically grounded intelligent systems.</p> |
| title | Recognizing the Unseen: A Verified, Multimodal Framework for Trauma-Informed AI |
| topic | trauma-informed AI multimodal emotion recognition affective computing symbolic verification Fuzzy Logic Z3 solver microexpression analysis ethical AI empathy modeling |
| url | https://doi.org/10.5281/zenodo.17705532 |