PIPBench: A Profile-Inclusive Framework for Personalized Image Generation Evaluation
Recent text-to-image models such as DALLE-3 excel at following diverse prompts yet remain blind to individual aesthetic preferences. We study personalized image generation, where models must align outputs with a user's implicit visual preferences based on a few historically preferred images and a short prompt. To this end, we introduce PIPBench, the first profile-inclusive benchmark for evaluating personalized image generation. We further propose a novel data construction pipeline that leverages psychological and demographic profiling dimensions for both real-user data collection and scalable
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
- Linked via arxiv authorYuhang Wu →
PIPBench: A Profile-Inclusive Framework for Personalized Image Generation Evaluation
- Linked via arxiv authorShuxiang Zhang →
PIPBench: A Profile-Inclusive Framework for Personalized Image Generation Evaluation
- Linked via arxiv authorWee Hian Ching →
PIPBench: A Profile-Inclusive Framework for Personalized Image Generation Evaluation
- Linked via arxiv authorChi Zhang →
PIPBench: A Profile-Inclusive Framework for Personalized Image Generation Evaluation
- Linked via arxiv authorMiao Liu →
PIPBench: A Profile-Inclusive Framework for Personalized Image Generation Evaluation