AI Paper Atlas

Research map / 3D vision / Face and shape

SRAM: Shape-Realism Alignment Metric for No Reference 3D Shape Evaluation

Sheng Liu, Tianyu Luan, Phani Nuney, Xuelu Feng, Junsong Yuan

AAAI 2026 · 3D vision › Face and shape · approach: Optimization-based shape matching · 1 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
use SRAM LLM-bridge metric instead of PointNet ad-hoc baselineno-reference realism metric · RealismGrading (k-fold over objects)SROCC: 0.591 -> 0.696Improved
full finetune the LLM instead of LoRA r=4 (ablation)SRAM · RealismGrading (k-fold over objects)PLCC: 0.656 -> 0.689Improved
use full finetuning instead of prefix tuning (ablation)SRAM · RealismGrading (k-fold over objects)PLCC: 0.081 -> 0.689Improved
use smaller LoRA rank (2 or 4 vs 8 or 16) (ablation)SRAM with LoRA · RealismGrading (k-fold over objects)PLCC: 0.550 (r=8) -> 0.656 (r=4)Improved
regress score with realism decoder instead of generating text tokens (ablation)SRAM · RealismGrading (k-fold over objects)PLCC: 0.245 -> 0.689Improved
omit object name from realism prompt (ablation)SRAM · RealismGrading (k-fold over objects)PLCC: 0.567 -> 0.689Improved

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