AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Scaling Up Personalized Image Aesthetic Assessment via Task Vector Customization

Jooyeol Yun, Jaegul Choo

ECCV 2024 · Image generation › Autoregressive and latent generation · approach: Learning-based image quality assessment · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
customize layer-wise task vectors from GIAA/IQA databases with trainable coefficientsPIAA with OpenCLIP ViT-L/14 · REAL-CUR cross-database 10-shotSROCC: 0.495 -> 0.577 (vs MTCL)Improved
use task vector method cross-database vs models trained on AADBPIAA task vector customization · AADB 100-shotSROCC: 0.622 -> 0.654 (MTCL trained on AADB vs ours)Improved
task vector customization vs best prior method in-domainPIAA · Flickr-AES 10-shotSROCC: 0.667 -> 0.668 (MTCL vs ours)No clear effect
keep task vectors frozen and train only coefficients instead of fine-tuning the best-fit model (ablation)PIAA ViT-L/14 · REAL-CUR 10-shotSROCC: 0.303 -> 0.577Improved
use layer-wise task vectors rather than one per task (ablation)PIAA ViT-L/14 · REAL-CUR 10-shotSROCC: 0.438 -> 0.572Improved
use Bradley-Terry rank loss instead of MSE loss (ablation)PIAA ViT-L/14 · REAL-CUR 10-shotSROCC: 0.553 -> 0.572Improved
adaptive softmax-SROCC coefficient initialization instead of uniform (ablation)PIAA ViT-L/14 · REAL-CUR 10-shotSROCC: 0.520 -> 0.577Improved
adaptive initialization instead of best-fit initialization (ablation)PIAA ViT-L/14 · REAL-CUR 10-shotSROCC: 0.464 -> 0.577Improved
increase the number of databases (task vectors) usedPIAA ViT-L/14 · REAL-CURSROCCImproved
use smaller ViT-B/16 backbone instead of ViT-L/14 (ablation)PIAA task vectors · REAL-CUR 10-shotSROCC: 0.577 -> 0.562; FPS 94.41 -> 417.95Mixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy