arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2310.02422cs.LGcs.AIcs.DCcs.MMcs.NI
收藏

OneAdapt：通过反向传播实现视频分析应用的快速配置自适应

OneAdapt: Fast Configuration Adaptation for Video Analytics Applications via Backpropagation

University of Chicago（芝加哥大学）；Stanford University（斯坦福大学）；Microsoft Research（微软研究院）

Kuntai Du, Yuhan Liu, Yitian Hao, Qizheng Zhang, Haodong Wang, Yuyang Huang, Ganesh Ananthanarayanan, Junchen Jiang

备注：
SoCC' 23
阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对视频分析应用现有配置自适应技术的不足，提出OneAdapt框架，利用DNN可微性计算AccGrad实现梯度上升式配置适配，在多场景下显著降低资源开销或提升推理精度。

摘要：

针对流媒体数据的深度学习推理（如视频或激光雷达数据流中的目标检测、音频波形中的文本提取）如今已无处不在。为实现高推理精度，这类应用通常需要大量网络带宽来采集高保真数据，还需要充足的GPU资源来运行深度神经网络（DNN）。虽然通过最优适配视频分辨率、帧率等配置旋钮可大幅降低对网络带宽和GPU资源的高需求，但现有自适应技术无法同时满足三项要求：（i）以最小的额外GPU或带宽开销适配配置；（ii）基于数据对最终DNN精度的影响做出接近最优的决策；（iii）支持多种配置旋钮的自适应。本文提出OneAdapt，它采用梯度上升策略适配配置旋钮，从而满足上述所有要求。其核心思路是利用DNN的可微性，快速估计精度相对于每个配置旋钮的梯度（称为AccGrad）。具体而言，OneAdapt通过将两个梯度相乘来估计AccGrad：InputGrad（即每个配置旋钮如何影响DNN的输入）和DNNGrad（即DNN输入如何影响DNN推理输出）。我们在五种配置类型、四项分析任务和五种输入数据上对OneAdapt进行了评估。与最先进的自适应方案相比，OneAdapt在保持相当精度的同时，将带宽使用量和GPU使用量降低了15-59%；或在使用等量或更少资源的情况下，将精度提升了1-5%。

英文摘要：

Deep learning inference on streaming media data, such as object detection in video or LiDAR feeds and text extraction from audio waves, is now ubiquitous. To achieve high inference accuracy, these applications typically require significant network bandwidth to gather high-fidelity data and extensive GPU resources to run deep neural networks (DNNs). While the high demand for network bandwidth and GPU resources could be substantially reduced by optimally adapting the configuration knobs, such as video resolution and frame rate, current adaptation techniques fail to meet three requirements simultaneously: adapt configurations (i) with minimum extra GPU or bandwidth overhead; (ii) to reach near-optimal decisions based on how the data affects the final DNN's accuracy, and (iii) do so for a range of configuration knobs. This paper presents OneAdapt, which meets these requirements by leveraging a gradient-ascent strategy to adapt configuration knobs. The key idea is to embrace DNNs' differentiability to quickly estimate the accuracy's gradient to each configuration knob, called AccGrad. Specifically, OneAdapt estimates AccGrad by multiplying two gradients: InputGrad (i.e. how each configuration knob affects the input to the DNN) and DNNGrad (i.e. how the DNN input affects the DNN inference output). We evaluate OneAdapt across five types of configurations, four analytic tasks, and five types of input data. Compared to state-of-the-art adaptation schemes, OneAdapt cuts bandwidth usage and GPU usage by 15-59% while maintaining comparable accuracy or improves accuracy by 1-5% while using equal or fewer resources.

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。

↑