TraMP-LLaMA combines RGB and landmark trajectory inputs with decoupled instruction tuning to jointly output severity scores and structured motion reports, improving Spearman's correlation by at least 4.39% over baselines on an extended PFED5 dataset.
QAFE- Net: Quality assessment of facial expressions with landmark heatmaps
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
TraMP-LLaMA: Generative Interpretability with Decoupled Instruction Tuning for Facial Expression Quality Assessment
TraMP-LLaMA combines RGB and landmark trajectory inputs with decoupled instruction tuning to jointly output severity scores and structured motion reports, improving Spearman's correlation by at least 4.39% over baselines on an extended PFED5 dataset.