Deep Residual Learning for Image Recognition Supplementary Materials

In this section we introduce our detection method based on the baseline Faster R-CNN [6] system. The models are initialized by the ImageNet classification models, and then fine-tuned on the object detection data. We have experimented with ResNet-50/101 at the time of the ILSVRC & COCO 2015 detection competitions. Unlike VGG-16 used in [6], our ResNet has no hidden fc layers. We adopt the idea of “Networks on Conv feature maps” (NoC) [7] to address this issue. We compute the full-image shared conv feature maps using those layers whose strides on the image are no greater than 16 pixels (i.e., conv1, conv2 x, conv3 x, and conv4 x, totally 91 conv layers in ResNet-101). We consider these layers as analogous to the 13 conv layers in VGG-16, and by doing so, both ResNet and VGG-16 have conv feature maps of the same total stride (16 pixels). These layers are shared by a region proposal network (RPN, generating 300 proposals) [6] and a Fast R-CNN detection network [2]. RoI pooling [2] is performed before conv5 1. On this RoI-pooled feature, all layers of conv5 x and up are adopted for each region, playing the roles of VGG-16’s fc layers. The final classification layer is replaced by two sibling layers (classification and box regression [2]). For the usage of BN layers, after pre-training, we compute the BN statistics (means and variances) for each layer on the ImageNet training set. Then the BN layers are fixed during fine-tuning for object detection. As such, the BN layers become linear activations with constant offsets and scales, and BN statistics are not updated by fine-tuning. We fix the BN layers mainly for reducing memory consumption in Faster R-CNN training.