SVGauge: Towards Human-Aligned Evaluation for SVG Generation
Not on the map. Why?
Abstract (arXiv)
Generated Scalable Vector Graphics (SVG) images demand evaluation criteria tuned to their symbolic and vectorial nature: criteria that existing metrics such as FID, LPIPS, or CLIPScore fail to satisfy. In this paper, we introduce SVGauge, the first human-aligned, reference based metric for text-to-SVG generation. SVGauge jointly measures (i) visual fidelity, obtained by extracting SigLIP image embeddings and refining them with PCA and whitening for domain alignment, and (ii) semantic consistency, captured by comparing BLIP-2-generated captions of the SVGs against the original prompts in the combined space of SBERT and TF-IDF. Evaluation on the proposed SHE benchmark shows that SVGauge attains the highest correlation with human judgments and reproduces system-level rankings of eight zero-shot LLM-based generators more faithfully than existing metrics. Our results highlight the necessity of vector-specific evaluation and provide a practical tool for benchmarking future text-to-SVG generation models.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 21 of its 26 references that are in the map. First posted on arXiv 8 Sep 2025.
Builds on
21 of 26 references.
References · 21
- 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
- 2023GPT-4 Technical Reportcited by 18,396
- 2024The Llama 3 Herd of Modelscited by 14,715
- 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
- 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,170
- 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,974
- 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
- 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,330
- 2023BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Modelscited by 6,137
- 2023Sigmoid Loss for Language Image Pre-Trainingcited by 3,109
- 2021CLIPScore: A Reference-free Evaluation Metric for Image Captioningcited by 2,424
- 2017Fine-tuning CNN Image Retrieval with No Human Annotationcited by 499
9 more
- 2023Florence-2: Advancing a Unified Representation for a Variety of Vision Taskscited by 410
- 2024Building and better understanding vision-language models: insights and future directionscited by 173
- 2022VectorFusion: Text-to-SVG by Abstracting Pixel-Based Diffusion Modelscited by 112
- 2023StarVector: Generating Scalable Vector Graphics Code from Images and Textcited by 102
- 2022Towards Layer-wise Image Vectorizationcited by 79
- 2023SVGDreamer: Text Guided SVG Generation with Diffusion Modelcited by 76
- 2023IconShop: Text-Guided Vector Icon Synthesis with Autoregressive Transformerscited by 75
- 2024Empowering LLMs to Understand and Generate Complex Vector Graphicscited by 66
- 2024Chat2SVG: Vector Graphics Generation with Large Language Models and Image Diffusion Modelscited by 38
Built on by
Citing papers per year: 2026: 1.
1 paper.
Similar, not cited
Close in wording, no citation either way
- 2023Beyond Pixels: Exploring Human-Readable SVG Generation for Simple Images with Vision Language Models
- 2025SVGThinker: Instruction-Aligned and Reasoning-Driven Text-to-SVG Generation
- 2025SVGenius: Benchmarking LLMs in SVG Understanding, Editing and Generation
- 2025DuetSVG: Unified Multimodal SVG Generation with Internal Visual Guidance
- 2025SVGen: Interpretable Vector Graphics Generation with Large Language Models
- 2025UniSVG: A Unified Dataset for Vector Graphic Understanding and Generation with Multimodal Large Language Models
- 2026IntroSVG: Learning from Rendering Feedback for Text-to-SVG Generation via an Introspective Generator-Critic Framework
- 2026VectorGym: A Multi-Task Benchmark for SVG Code Generation, Sketching and Editing