通过对比学习增强归一化流用于异常分割
FlowCLAS: Enhancing Normalizing Flow Via Contrastive Learning For Anomaly Segmentation
AI总结：
FlowCLAS通过对比学习增强归一化流，有效提升异常分割性能，达到多个机器人异常分割基准的最先进水平。
摘要：
异常分割是安全关键型机器人应用中至关重要的能力，必须能够识别意外事件。归一化流（NFs）是一类生成模型，由于其能够高效建模内群数据分布，成为该任务的有希望方法。然而，在动态场景中，复杂多模态的数据分布导致它们难以识别分布外样本，从而在性能上落后于领先判别方法。为了解决这一限制，我们引入了FlowCLAS，一种混合框架，通过对比损失增强传统最大似然目标。利用异常暴露，该目标显式地在潜在空间中强制正常和异常特征的分离，保留了NFs的概率基础，同时嵌入了它们所缺乏的判别能力。该方法的强大力量通过FlowCLAS在多个具有挑战性的异常分割基准测试中建立新的最先进（SOTA）性能得到证明，包括Fishyscapes Lost & Found、Road Anomaly、SegmentMeIfYouCan-ObstacleTrack和ALLO。我们的实验还显示，这种对比方法比其他基于异常的训练策略更有效，成功弥合了性能差距，使NFs的性能接近领先判别方法。
英文摘要：
Anomaly segmentation is an essential capability for safety-critical robotics applications that must be aware of unexpected events. Normalizing flows (NFs), a class of generative models, are a promising approach for this task due to their ability to model the inlier data distribution efficiently. However, their performance falters in dynamic scenes, where complex, multi-modal data distributions cause them to struggle with identifying out-of-distribution samples, leaving a performance gap to leading discriminative methods. To address this limitation, we introduce FlowCLAS, a hybrid framework that enhances the traditional maximum likelihood objective of NFs with a discriminative, contrastive loss. Leveraging Outlier Exposure, this objective explicitly enforces a separation between normal and anomalous features in the latent space, retaining the probabilistic foundation of NFs while embedding the discriminative power they lack. The strength of this approach is demonstrated by FlowCLAS establishing new state-of-the-art (SOTA) performance across multiple challenging anomaly segmentation benchmarks for robotics, including Fishyscapes Lost & Found, Road Anomaly, SegmentMeIfYouCan-ObstacleTrack, and ALLO. Our experiments also show that this contrastive approach is more effective than other outlier-based training strategies for NFs, successfully bridging the performance gap to leading discriminative methods. Project page: https://trailab.github.io/FlowCLAS
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
