We address the challenge of training a large supernet for the object detection task, using a relatively small amount of training data. Specifically, we propose an efficient supernet-based neural architecture search (NAS) method that uses transfer learning and search space pruning. First, the supernet is pre-trained on a classification task, for which large datasets are available. Second, the search space defined by the supernet is pruned by removing candidate models that are predicted to perform poorly. To effectively remove the candidates over a wide range of resource constraints, we particularly design a performance predictor, called path filter, which can accurately predict the relative performance of the models that satisfy similar resource constraints. Hence, supernet training is more focused on the best-performing candidates. Our path filter handles prediction for paths with different resource budgets. Compared to once-for-all, our proposed method reduces the computational cost of the optimal network architecture by 30% and 63%, while yielding better accuracy-floating point operations Pareto front (0.85 and 0.45 points of improvement on average precision for Pascal VOC and COCO, respectively).

翻译：针对使用相对少量训练数据训练大型超网络用于目标检测任务的挑战，我们提出了一种基于超网络的高效神经架构搜索（NAS）方法，该方法融合了迁移学习与搜索空间剪枝技术。首先，超网络在分类任务上进行预训练（该领域存在大规模数据集）；其次，通过移除被预测性能不佳的候选模型对超网络定义的搜索空间进行剪枝。为有效移除不同资源约束条件下的候选模型，我们专门设计了名为路径过滤器的性能预测器，能够准确预测满足相似资源约束模型的相对性能，从而使超网络训练更聚焦于最优候选模型。该路径过滤器可处理不同资源预算下路径的预测任务。与Once-for-All方法相比，本方法在降低最优网络架构计算成本30%与63%的同时，实现了更优的精度-浮点运算Pareto前沿（在Pascal VOC和COCO数据集上平均精度分别提升0.85和0.45个百分点）。