For a long time, I've been rethinking AIFFD's positioning—it shouldn't just be a style assessment tool, but a research-driven intelligent system for personal style.
The core logic behind AIFFD is gradually becoming clear:
Can personal style be presented as an evolving profile, rather than a fixed label?
AIFFD currently explores how to combine multi-layered information, including body structure, facial features, color characteristics, personal preferences, and real-world feedback, to construct a continuously evolving personal style profile.
This also prompted me to clarify the relationship between the two parts of the project:
AIFFD is the functional product and research prototype.
It includes assessment, user profiling, recommendation logic, feedback, and future product validation.
OhSammi Research is the research information behind the system.
I need to document the issues, methodologies, design decisions, system iterations, and broader research topics that have arisen during the development of AIFFD.
I am clear about the areas that need in-depth exploration:
AI-assisted visual recommendations
Personalized female identity and self-presentation
Trust and explainability
User autonomy in AI-assisted decision-making
As my research deepens, I find this increasingly important: personal style should not be viewed as a static category.
An initial assessment might generate a style profile V1.0, but future outfit feedback, preferences, purchasing decisions, and adjustments can progressively update this profile.
This shift from fixed style tags to continuously evolving personal style profiles is becoming one of AIFFD's core principles.
The system is still under development, and I will continue to document the research issues, decisions, and lessons learned along the way.
Explore AIFFD: aiffd.com
Research Log: en.ohsammi.com