This paper proposes anchor pruning for object detection in one-stage anchor-based detectors. While pruning techniques are widely used to reduce the computational cost of convolutional neural networks, they tend to focus on optimizing the backbone networks where often most computations are. In this work we demonstrate an additional pruning technique, specifically for object detection: anchor pruning. With more efficient backbone networks and a growing trend of deploying object detectors on embedded systems where post-processing steps such as non-maximum suppression can be a bottleneck, the impact of the anchors used in the detection head is becoming increasingly more important. In this work, we show that many anchors in the object detection head can be removed without any loss in accuracy. With additional retraining, anchor pruning can even lead to improved accuracy. Extensive experiments on SSD and MS COCO show that the detection head can be made up to 44% more efficient while simultaneously increasing accuracy. Further experiments on RetinaNet and PASCAL VOC show the general effectiveness of our approach. We also introduce `overanchorized' models that can be used together with anchor pruning to eliminate hyperparameters related to the initial shape of anchors.
翻译:本文提议在以锚为主的单级探测器中进行物体探测的锚线修剪。 虽然修剪技术被广泛用于降低进化神经网络的计算成本, 但它们往往侧重于优化主干网, 通常大多数计算都是这样的。 在这项工作中, 我们展示了一种额外的修剪技术, 特别是用于物体探测的修剪技术: 锚皮切割。 随着更高效的主干网络和在嵌入系统上部署物体探测器的日益增长的趋势, 在嵌入系统中, 非最大抑制等后处理步骤可能是一个瓶颈, 探测头中使用的锚的影响越来越重要。 在这项工作中, 我们显示许多物体探测头中的锚可以在不丢失任何准确性的情况下被移除。 通过进一步的再培训, 锚线划甚至可以提高准确性。 SSD 和 MS COCO 的广泛实验表明, 探测头在提高44% 的效率的同时, 提高精确性。 在Retinnet 和 PCAL VOC 等后处理步骤的进一步实验显示了我们的方法的总体有效性。 我们还引入了“ 超紧” 模型, 与锚钻头有关的固定形状相关。