YOLO 真的需要在每个 epoch 看到每一张训练图像吗？
Does YOLO Really Need to See Every Training Image in Every Epoch?
AFSS 把「每张图的学习充分度」定义为检测召回率与精确率的最小值，并据此动态划分易/中/难样本、按连续复习方式稀疏重采样，可直接迁移到任意检测器或分割模型的训练循环中，作为降低每轮数据量、加速收敛的通用数据调度手段。
原文依据 AFSS achieves more than 1.43x training speedup for YOLO-series detectors while also improving accuracy
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论YOLO 检测器以推理速度快著称，但其训练过程却出人意料地耗时，原因在于其穷举式流程会在每个 epoch 处理每一张训练图像，即使许多图像早已被充分学习。这与「You Only Look Once」理念所暗示的效率形成鲜明反差。这自然引出一个重要问题：YOLO 真的需要在每个 epoch 看到每一张训练图像吗？为探究这一问题，我们提出一种抗遗忘采样策略（Anti-Forgetting Sampling Strategy，AFSS），它在每个 epoch 动态决定哪些图像应被使用、哪些可以被跳过，使检测器能够更有效且更高效地学习。具体而言，AFSS 以每张训练图像的检测召回率与精确率的最小值来衡量其学习充分度，并据此将训练图像动态划分为易、中、难三个等级。易训练图像在训练中以连续复习的方式被稀疏重采样，并优先选择长时间未被使用的图像，以减少冗余并防止遗忘。中等训练图像被部分选取，优先选择近期未使用的图像，其余则从尚未选中的图像中随机抽取，以保证覆盖并防止遗忘。难训练图像在每个 epoch 被全部采样，以确保充分学习。每张训练图像的学习充分度会被定期更新，使检测器能够随时间自适应地将注意力转向信息量大的训练图像，同时逐步丢弃冗余图像。在广泛使用的自然图像检测基准（MS COCO 2017 与 PASCAL VOC 2007）以及遥感检测数据集（DOTA-v1.0 与 DIOR-R）上，AFSS 为 YOLO 系列检测器带来超过 1.43 倍的训练加速，同时提升了精度。
英文原文摘要 1863 字符
YOLO detectors are known for their fast inference speed, yet training them remains unexpectedly time-consuming due to their exhaustive pipeline that processes every training image in every epoch, even when many images have already been sufficiently learned. This stands in clear contrast to the efficiency suggested by the "You Only Look Once" philosophy. This naturally raises an important question: Does YOLO really need to see every training image in every epoch? To explore this, we propose an Anti-Forgetting Sampling Strategy (AFSS) that dynamically determines which images should be used and which can be skipped during each epoch, allowing the detector to learn more effectively and efficiently. Specifically, AFSS measures the learning sufficiency of each training image as the minimum of its detection recall and precision, and dynamically categorizes training images into easy, medium, or hard levels accordingly. Easy training images are sparsely resampled during training in a continuous review manner, with priority given to those that have not been used for a long time to reduce redundancy and prevent forgetting. Moderate training images are partially selected, prioritizing recently unused ones and randomly choosing the rest from unselected images to ensure coverage and prevent forgetting. Hard training images are fully sampled in every epoch to ensure sufficient learning. The learning sufficiency of each training image is periodically updated, enabling detectors to adaptively shift its focus toward the informative training images over time while progressively discarding redundant ones. On widely used natural image detection benchmarks (MS COCO 2017 and PASCAL VOC 2007) and remote sensing detection datasets (DOTA-v1.0 and DIOR-R), AFSS achieves more than 1.43x training speedup for YOLO-series detectors while also improving accuracy.