Split Computing for Complex Object Detectors: Challenges and Preliminary Results
Split Computing for Complex Object Detectors: Challenges and Preliminary Results
复制标题
复杂目标检测器的分割计算：挑战和初步结果
DOI：
10.1145/3410338.3412338
复制
发表时间：
2020
期刊：
影响因子：
--
通讯作者：
M. Levorato
中科院分区：
文献类型：
--
作者：
Yoshitomo Matsubara;M. Levorato
随着深度神经网络模型的移动计算和边缘计算的趋势，一种中间选择——分裂计算已经引起了研究界的关注。以往的研究经验表明，虽然移动和边缘计算通常是总推理时间方面的最佳选择，但在某些情况下，分裂计算方法可以实现更短的推理时间。然而，所有提出的分割计算方法都集中在图像分类任务上，并且大多数都是在远离实际场景的小数据集上进行评估的。在本文中，我们讨论了为在大型数据集COCO 2017上训练的强大R-CNN目标检测器开发拆分计算方法所面临的挑战。我们从分层张量大小和模型大小的角度对目标检测器进行了广泛的分析，并表明朴素的分割计算方法不会减少推理时间。据我们所知，这是第一个为这种物体探测器注入小瓶颈的研究，并揭示了分裂计算方法的潜力。
Following the trends of mobile and edge computing for DNN models, an intermediate option, split computing, has been attracting attentions from the research community. Previous studies empirically showed that while mobile and edge computing often would be the best options in terms of total inference time, there are some scenarios where split computing methods can achieve shorter inference time. All the proposed split computing approaches, however, focus on image classification tasks, and most are assessed with small datasets that are far from the practical scenarios. In this paper, we discuss the challenges in developing split computing methods for powerful R-CNN object detectors trained on a large dataset, COCO 2017. We extensively analyze the object detectors in terms of layer-wise tensor size and model size, and show that naive split computing methods would not reduce inference time. To the best of our knowledge, this is the first study to inject small bottlenecks to such object detectors and unveil the potential of a split computing approach.
DOI：
10.1109/icpr48806.2021.9412388
发表时间：
2020-07
期刊：
2020 25th International Conference on Pattern Recognition (ICPR)
影响因子：
--
作者：
Yoshitomo Matsubara;M. Levorato
通讯作者：
Yoshitomo Matsubara;M. Levorato