English(EN) FORGE: Forward-Only Test-Time Adaptation for Integer-Only Vision Models on Microcontrollers

新的FORGE方法实现了面向微控制器上纯整数视觉模型的测试时自适应

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-03 04:00

研究人员开发了FORGE，这是一种新颖的正向测试时自适应方法，专门为在微控制器上运行的纯整数视觉模型设计。该方法通过仅使用前向传播估计来重新归一化折叠的卷积层，解决了在没有通常所需的反向传播机制的情况下将模型适应现实世界分布偏移的挑战。FORGE展示了显著的准确性提升，恢复了基于梯度的方法的大部分优势，并且足够高效，可以以最小的能源和时间成本部署在ESP32-S3等微控制器上。 AI

影响使得在资源受限的微控制器上部署更鲁棒的AI模型成为可能，提高了在现实条件下的性能。

排序理由该集群包含一篇详细介绍新AI模型自适应方法的论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.LG 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 22 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群包含一篇详细介绍新AI模型自适应方法的论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, infra; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.LG TIER_1 English(EN) · Muhammad Rehan, Haider Ali, Muhammad Ali Munir, Moaz Amjad · 2026-09-03 04:00

FORGE：面向微控制器上仅整数视觉模型的正向仅测试时自适应

arXiv:2609.01683v1 Announce Type: cross Abstract: Vision models deployed on microcontrollers (MCUs) are quantized to integer-only arithmetic and run in inference-only runtimes that do not carry the machinery backpropagation needs: the standard tool for adapting a model to the dis…

本文如何被排名

报道来源 [1]

FORGE：面向微控制器上仅整数视觉模型的正向仅测试时自适应

相关实体

相关话题