arXiv cs.AI· Yicong Li, Junjie Wang, Leander Lauenburg, Ella Hugie, Alexandra Irger, Wanhua Li, Donglai Wei, Hanspeter Pfister·· 12 小时前AI 评分29
视觉语言模型在连接组学突触与神经元校对中的基准测试
Benchmarking Vision-Language Models on Synapse and Neuron Proofreading in Connectomics
AI 导读
研究对视觉语言模型在连接组学电子显微镜图像校对任务上进行了基准测试：突触校对（存在性与极性）评估了19个开源和2个闭源模型，神经元校对（分裂错误与合并错误）评估了3个开源和2个闭源模型。多数模型零样本表现仅相当于随机水平，少量示例主要帮助闭源和最大的开源模型；在数千个标签上做LoRA微调后，开源模型达到专家模型水平，且在未见物种上识别合并错误时优于专家模型。
正文 · 原文
Abstract:We benchmarked vision-language models (VLMs) on the decisions annotators take when inspecting electron microscopy images in connectomics: synapse proofreading (presence and polarity) and neuron proofreading (split errors and merge errors). For synapse proofreading, we evaluated 19 open and 2 closed models across various architectures and sizes under zero-shot, four-shot in-context learning and LoRA settings, against specialist models, on datasets constructed by us using public resources. For neuron proofreading, we evaluated 3 open and 2 closed models on the ConnectomeBench2 dataset, with cross-species transfer from fly and mouse to human and zebrafish. Most models were at chance zero-shot; a few examples helped mainly the closed and largest open ones. LoRA on a few thousand labels brought open models level with specialist models. When evaluated on unseen species, the best adapted VLMs outperformed specialist models in identifying merge errors. The project will be publicly available upon acceptance.
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2609.36492 [cs.CV] |
| (or arXiv:2609.36492v2 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2609.36492 arXiv-issued DOI via DataCite |
Submission history
From: Yicong Li [view email]
[v1]
Tue, 29 Sep 2026 01:48:42 UTC (561 KB)
[v2]
Fri, 23 Oct 2026 05:15:15 UTC (561 KB)
来源：arXiv cs.AI · arxiv.org