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 search space pruning. 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 for supernet, called path filter, which is conditioned by resource constraints and 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：针对目标检测任务中利用少量训练数据训练大型超网络的挑战，我们提出了一种基于搜索空间剪枝的高效超网络神经架构搜索（NAS）方法。通过移除预测性能较差的候选模型来剪枝超网络定义的搜索空间。为有效移除广泛资源约束下的候选模型，我们专门设计了面向超网络的性能预测器——路径过滤器，该过滤器以资源约束为条件，能精准预测满足相似资源约束条件下模型的相对性能，从而让超网络训练更聚焦于性能最优的候选模型。我们的路径过滤器可处理不同资源预算条件下路径的预测任务。与一次性训练方法相比，本方法在实现更优准确率-浮点运算次数帕累托前沿（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）的同时，最优网络架构的计算开销分别降低了30%和63%。