[论文解读] A Scalable Multilabel Classification to Deploy Deep Learning Architectures For Edge Devices

Tolulope A. Odetola, Ogheneuriri Oderhohwo+1 more|arXiv (Cornell University)|Nov 5, 2019

Advanced Image and Video Retrieval Techniques参考文献 24被引用 5

一句话总结

本文提出了一种可扩展的单模型多标签分类框架，适用于边缘设备部署，采用单一CNN模型结合多损失和多准确率层，无需图像裁剪或拼接处理。该方法在AlexNet、GoogleNet和SqueezeNet上实现与传统多标签方法相当的准确率，同时将MACC操作减少1.8倍，延迟降低最高达50%，并减小了模型尺寸。

ABSTRACT

Convolution Neural Networks (CNN) have performed well in many applications such as object detection, pattern recognition, video surveillance and so on. CNN carryout feature extraction on labelled data to perform classification. Multi-label classification assigns more than one label to a particular data sample in a data set. In multi-label classification, properties of a data point that are considered to be mutually exclusive are classified. However, existing multi-label classification requires some form of data pre-processing that involves image training data cropping or image tiling. The computation and memory requirement of these multi-label CNN models makes their deployment on edge devices challenging. In this paper, we propose a methodology that solves this problem by extending the capability of existing multi-label classification and provide models with lower latency that requires smaller memory size when deployed on edge devices. We make use of a single CNN model designed with multiple loss layers and multiple accuracy layers. This methodology is tested on state-of-the-art deep learning algorithms such as AlexNet, GoogleNet and SqueezeNet using the Stanford Cars Dataset and deployed on Raspberry Pi3. From the results the proposed methodology achieves comparable accuracy with 1.8x less MACC operation, 0.97x reduction in latency and 0.5x, 0.84x and 0.97x reduction in size for the generated AlexNet, GoogleNet and SqueezeNet CNN models respectively when compared to conventional ways of achieving multi-label classification like hard-coding multi-label instances into single labels. The methodology also yields CNN models that achieve 50\% less MACC operations, 50% reduction in latency and size of generated versions of AlexNet, GoogleNet and SqueezeNet respectively when compared to conventional ways using 2 different single-labelled models to achieve multi-label classification.

研究动机与目标

• 解决现有多标签CNN在边缘设备上因高内存和计算成本导致的可扩展性和部署挑战。
• 消除多标签分类中对图像预处理步骤（如裁剪或拼接）的需求。
• 使单一深度学习模型能够高效地对多个互斥标签属性进行分类。
• 与传统的多模型或硬编码单标签方法相比，降低推理延迟、MACC操作次数和模型尺寸。
• 利用迁移学习和统一架构，实现在Raspberry Pi 3等边缘硬件上高效、低成本的多标签训练与部署。

提出的方法

• 设计一个具有多个损失层和准确率层的单一CNN模型，每个层对应一个互斥的标签属性。
• 使用迁移学习微调最先进的模型（AlexNet、GoogleNet、SqueezeNet）以实现多标签分类，而无需对网络架构进行大规模修改。
• 使用统一损失函数联合所有标签特定损失进行训练，以同时优化多个输出。
• 应用数据融合技术，处理同一模型中相关与无关标签分类任务。
• 将训练好的模型部署到边缘设备（如Raspberry Pi 3）并评估其推理效率、延迟和模型尺寸。
• 使用Caffe框架进行模型训练，使用NetScope可视化最终模型架构。

实验结果

研究问题

• RQ1 是否能够通过具有多个损失和准确率层的单一CNN模型实现有效的多标签分类，而无需图像预处理（如裁剪或拼接）？
• RQ2 与使用硬编码多标签实例或独立单标签模型的传统方法相比，该方法在准确率、延迟和模型尺寸方面表现如何？
• RQ3 迁移学习与多损失训练在多大程度上可减少边缘部署的MACC操作次数和内存占用？
• RQ4 该方法是否在保持高分类准确率的同时，实现了在资源受限设备上的可扩展、低成本多标签推理？

主要发现

• 所提方法在实现与传统方法相当的多标签分类准确率的同时，平均将MACC操作减少1.8倍。
• 与硬编码多标签训练方法相比，延迟降低0.97倍（约3%的性能提升）。
• 与使用两个独立单标签模型进行多标签分类相比，模型尺寸平均减少50%。
• 与硬编码多标签训练相比，AlexNet、GoogleNet和SqueezeNet的模型尺寸分别减少0.5倍、0.84倍和0.97倍。
• 该方法可在Raspberry Pi 3等边缘设备上实现高效部署，显著降低计算与内存开销。
• 该框架通过数据融合支持相关与无关属性的同时分类，提升了可扩展性，且未增加模型复杂度。