AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

SVGEval: A Vision-Grounded Framework for Perceptual-Quality Benchmarking and Evaluation in Text-to-SVG Generation

Yiming Wang, Ye Chen, Hanqi Chen, Bingbing Ni

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Interleaved image-text generation with MLLMs · 0 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
add rationale (reasoning) supervision (ablation)SVGEval-Scorer · SVGEval Part IIMAE: 0.8330 -> 0.5772Improved
add rationale (reasoning) supervision (ablation)SVGEval-Scorer · SVGEval Part IIadjacent accuracy: 60.30 -> 71.95Improved
add SVG code input alongside image (ablation)SVGEval-Scorer · SVGEval Part IIMAE: 0.6220 -> 0.5772Improved
add SVG code input alongside image (ablation)SVGEval-Scorer · SVGEval Part II SpatialMAE: 0.6400 -> 0.5167Improved
add SVG code input alongside image (ablation)SVGEval-Scorer · SVGEval Part IIadjacent accuracy: 70.40 -> 71.95Improved
add rendered image input alongside SVG code (ablation)SVGEval-Scorer · SVGEval Part IIMAE: 0.9557 -> 0.5772Improved
add rendered image input alongside SVG code (ablation)SVGEval-Scorer · SVGEval Part IIadjacent accuracy: 57.75 -> 71.95Improved
fine-tune Qwen3-VL 30B A3B scorer vs best general MLLM (Gemini-3 Pro)SVGEval-Scorer · SVGEval Part IIMAE: 0.7068 -> 0.5772Improved
fine-tune scorer vs best general MLLM on Spatial dimensionSVGEval-Scorer · SVGEval Part II SpatialMAE: 0.8400 -> 0.5167Improved
fine-tune scorer vs Gemini-3 Pro on Semantic dimensionSVGEval-Scorer · SVGEval Part II SemanticMAE: 0.6160 -> 0.6129No clear effect
compare models on Spatial versus Semantic dimensionGPT-5 · SVGEval Part IIMAE: 0.7840 -> 1.2258Hurt

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