面向端到端自动驾驶系统的缩放感知数据选择
Scaling-Aware Data Selection for End-to-End Autonomous Driving Systems
为端到端自动驾驶规划器提供了一套按域拟合神经缩放定律、再迭代优化数据配比的数据选择流程，可直接迁移到视觉模型训练集构建与指标驱动的数据采集决策中。
原文依据 fitting neural scaling laws from each data domain to the evaluation metrics
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论面向物理人工智能应用的大规模深度学习模型依赖于多样化的训练数据采集工作。这些模型以及相应的训练数据，必须满足模型能够在真实环境中部署所需的各项不同评价标准。数据选择策略可以指导训练集的构建，但现有框架并未考虑数据点对不同指标影响的模糊性。在本工作中，我们提出基于缩放感知迭代采集的混合优化（MOSAIC），这是一个通用的数据选择框架，其运作方式为：(i) 将数据集划分为若干域；(ii) 为每个数据域拟合到评价指标的神经缩放定律；(iii) 通过迭代地加入那些能使指标变化最大化的域中的数据，来优化数据混合配比。我们将 MOSAIC 应用于自动驾驶（AD），其中端到端（E2E）规划器模型在扩展预测驾驶员模型得分（EPDMS）上进行评估，该得分是驾驶规则合规性指标的聚合。在此，MOSAIC 在 EPDMS 上优于多种基线，且所用数据最多减少 80%。
英文原文摘要 1129 字符
Large-scale deep learning models for physical AI applications depend on diverse training data collection efforts. These models and correspondingly, the training data, must address the different evaluation criteria necessary for the models to be deployable in real-world environments. Data selection policies can guide the development of the training set, but current frameworks do not account for the ambiguity in how data points affect different metrics. In this work, we propose Mixture Optimization via Scaling-Aware Iterative Collection (MOSAIC), a general data selection framework that operates by: (i) partitioning the dataset into domains; (ii) fitting neural scaling laws from each data domain to the evaluation metrics; and (iii) optimizing a data mixture by iteratively adding data from domains that maximize the change in metrics. We apply MOSAIC to autonomous driving (AD), where an End-to-End (E2E) planner model is evaluated on the Extended Predictive Driver Model Score (EPDMS), an aggregate of driving rule compliance metrics. Here, MOSAIC outperforms a diverse set of baselines on EPDMS with up to 80% less data.