用流匹配联合训练编码器与生成器，提升自监督学习效率与性能。
High-Performance Self-Supervised Learning by Joint Training of Flow Matching
- 通过解耦编码器与条件流匹配生成器，实现高质量生成与有效识别并行。
- 在可穿戴数据上训练时间减少50.4%，推理速度提升最高51.0倍。
- 适合追求高效高质自监督学习的工业与边缘智能应用。
扩散模型在数据生成过程中可学习丰富表征，具备自监督学习潜力，但存在生成质量与判别性能的权衡，且迭代采样带来高昂计算与能耗，限制其在工业和边缘AI中的应用。为此，我们提出基于流匹配的基础模型FlowFM，联合训练表示编码器与条件流匹配生成器。该解耦设计同时实现高保真生成与有效识别。通过使用流匹配学习更简单的速度场，FlowFM加速并稳定了训练过程，提升了表征学习效率。在可穿戴传感器数据上的实验表明，与基于扩散的方法相比，FlowFM将训练时间减少50.4%。在下游任务中，FlowFM在全部五个数据集上超越当前最优自监督方法（SSL-Wearables），最高实现51.0倍的推理加速，同时保持高生成质量。代码已开源：https://github.com/Okita-Laboratory/jointOptimizationFlowMatching。
原文摘要 · Abstract (English)
Diffusion models can learn rich representations during data generation, showing potential for Self-Supervised Learning (SSL), but they face a trade-off between generative quality and discriminative performance. Their iterative sampling also incurs substantial computational and energy costs, hindering industrial and edge AI applications. To address these issues, we propose the Flow Matching-based Foundation Model (FlowFM), which jointly trains a representation encoder and a conditional flow matching generator. This decoupled design achieves both high-fidelity generation and effective recognition. By using flow matching to learn a simpler velocity field, FlowFM accelerates and stabilizes training, improving its efficiency for representation learning. Experiments on wearable sensor data show FlowFM reduces training time by 50.4\% compared to a diffusion-based approach. On downstream tasks, FlowFM surpassed the state-of-the-art SSL method (SSL-Wearables) on all five datasets while achieving up to a 51.0x inference speedup and maintaining high generative quality. The implementation code is available at https://github.com/Okita-Laboratory/jointOptimizationFlowMatching.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。