← 研踪 PaperTrailarXiv:2602.00249 · v2
SANEval: Open-Vocabulary Compositional Benchmarks with Failure-mode Diagnosis
首次提交 2026-01-31 03:10
版本更新 2026-11-09 08:10
雷达观测 2026-11-10 13:19
摘要
Text-to-image (T2I) models struggle with compositional prompts requiring multiple objects, bound attributes, accurate counts, or spatial relations. Existing benchmarks face limits from fixed detector vocabularies, aggregated scores, and varying human-validation protocols. We introduce SANEval (Spatial, Attribute, and Numeracy Evaluation), an open-vocabulary compositional benchmark with two key contributions: (i) a public dataset of ~5,250 prompts paired with images from six state-of-the-art T2I models, split into algorithmically generated Simple and fully human-written Hard partitions across five compositional tasks, and (ii) a modular evaluation pipeline combining an LLM prompt parser, an open-vocabulary detector with LLM synonym expansion, and scorers for attribute binding, spatial relations, and numeracy. Each scorer emits numerical scores and diagnostic feedback for missing, extra, or mis-bound elements. Validated against human ratings (n = 500), SANEval achieves positive Pearson correlations across all categories (overall r = 0.222), outperforming CompBench++ (-0.016) and GenEval (-0.030). A manual audit of 24,012 detections shows 96.58% precision, with ablation confirming that synonym expansion is critical to pipeline performance. Dataset: [https://huggingface.co/datasets/saneval/saneval-release](https://huggingface.co/datasets/saneval/saneval-release). Pipeline: [https://github.com/nielseni6/saneval](https://github.com/nielseni6/saneval).
作者备注：Accepted in NeurIPS 2026, E&D Track, Sydney
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v2 · 2026-11-09 08:10 · 最新观测版本
SANEval: Open-Vocabulary Compositional Benchmarks with Failure-mode Diagnosis
Text-to-image (T2I) models struggle with compositional prompts requiring multiple objects, bound attributes, accurate counts, or spatial relations. Existing benchmarks face limits from fixed detector vocabularies, aggregated scores, and varying human-validation protocols. We introduce SANEval (Spatial, Attribute, and Numeracy Evaluation), an open-vocabulary compositional benchmark with two key contributions: (i) a public dataset of ~5,250 prompts paired with images from six state-of-the-art T2I models, split into algorithmically generated Simple and fully human-written Hard partitions across five compositional tasks, and (ii) a modular evaluation pipeline combining an LLM prompt parser, an open-vocabulary detector with LLM synonym expansion, and scorers for attribute binding, spatial relations, and numeracy. Each scorer emits numerical scores and diagnostic feedback for missing, extra, or mis-bound elements. Validated against human ratings (n = 500), SANEval achieves positive Pearson correlations across all categories (overall r = 0.222), outperforming CompBench++ (-0.016) and GenEval (-0.030). A manual audit of 24,012 detections shows 96.58% precision, with ablation confirming that synonym expansion is critical to pipeline performance. Dataset: [https://huggingface.co/datasets/saneval/saneval-release](https://huggingface.co/datasets/saneval/saneval-release). Pipeline: [https://github.com/nielseni6/saneval](https://github.com/nielseni6/saneval).
查看此版本解读 →查看此版本原文 ↗