发表机构
Albumentations LLC（Albumentations有限责任公司）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究比较了五个库的七条RGB图像输入流水线，在固定设置下测量吞吐量和GPU内存，发现DALI最快，AlbumentationsX在多数配方上优于其他库，并报告了变换覆盖情况。
AI 中文摘要
图像增强流水线必须在模型使用批次之前交付完整的批次。我们比较了来自五个库的七条输入路径，从RGB JPEG文件开始，以同步的CUDA float16批次结束。我们手动匹配了各库之间的变换配方和参数，以使工作负载尽可能具有可比性。实验使用了57个选定的配方，批次大小为256，以及一台NVIDIA L4机器。吞吐量和峰值进程GPU内存同时记录在759次测量中。在所有路径共享的11个配方上，DALI和AlbumentationsX的中位吞吐量分别为5,029和4,679图像/秒，中位峰值GPU内存分别为2,086和1,852 MiB。更广泛的成对比较显示，AlbumentationsX在26/26个TorchVision配方、50/51个Kornia配方和25/26个Pillow配方上表现更优。DALI在所有22个共享配方上比AlbumentationsX更快，中位吞吐量比率为1.18倍。一项单独的普查报告了所选AlbumentationsX RGB目录中118个条目的覆盖情况。该研究在固定设置下测量输入准备；它不测量模型训练、数值等价性或每个库的最佳可达配置。基准代码：此https URL。
英文摘要
An image-augmentation pipeline must deliver a complete batch before a model can use it. We compare seven input paths from five libraries, starting with RGB JPEG files and ending with a synchronized CUDA float16 batch. We manually matched transformation recipes and parameters across libraries to make the workloads as comparable as possible. The experiment uses 57 selected recipes, a batch size of 256, and one NVIDIA L4 machine. Throughput and peak process GPU memory are recorded together in 759 measurements. On the 11 recipes shared by all paths, DALI and AlbumentationsX have median throughputs of 5,029 and 4,679 images/s, with median peak GPU memory of 2,086 and 1,852 MiB. Broader pairwise comparisons favor AlbumentationsX on 26/26 TorchVision recipes, 50/51 Kornia recipes, and 25/26 Pillow recipes. DALI is faster than AlbumentationsX on all 22 shared recipes, with a median throughput ratio of 1.18x. A separate census reports coverage of the 118 entries in a selected AlbumentationsX RGB catalog. The study measures input preparation at fixed settings; it does not measure model training, numerical equivalence, or the best attainable configuration of each library. Benchmark code: https://github.com/albumentations-team/benchmark.
Comments26 pages, 8 figures. Benchmark code: https://github.com/albumentations-team/benchmark
