面向军事目标识别的DRFCN深度网络设计及实现

自动目标识别（ATR）技术一直是军事领域中急需解决的重点和难点。本文设计并实现了一种新的面向军事目标识别应用的DRFCN深度网络。首先，在DRPN部分通过卷积模块稠密连接的方式，复用深度网络模型中每一层的特征，实现高质量的目标采样区域提取；其次，在DFCN部分通过融合高低层次特征图语义特征信息，实现采样区域目标类别和位置信息的预测；最后，给出了DRFCN深度网络模型结构以及参数训练方法。与此同时，进一步对DRFCN算法开展了实验分析与讨论：1）基于PASCAL VOC数据及进行对比试验，结果表明，由于卷积模块稠密连接的犯法，在目标识别平均准确率、实时性和深度网络模型大小方面，DRFCN算法均明显优于依有据与深度学习的目标识别算法；同事，验证了DRFCN算法可以有效解决梯度弥散和梯度膨胀问题。2）利用自建军事目标数据集进行实验，结果表明，DRFCN算法在准确率和实时性上满足军事目标识别任务。

SSD is an unified framework for object detection with a single network. It has been originally introduced in this research article.

This repository contains a TensorFlow re-implementation of the original Caffe code. At present, it only implements VGG-based SSD networks (with 300 and 512 inputs), but the architecture of the project is modular, and should make easy the implementation and training of other SSD variants (ResNet or Inception based for instance). Present TF checkpoints have been directly converted from SSD Caffe models.

The organisation is inspired by the TF-Slim models repository containing the implementation of popular architectures (ResNet, Inception and VGG). Hence, it is separated in three main parts:

The SSD Notebook contains a minimal example of the SSD TensorFlow pipeline. Shortly, the detection is made of two main steps: running the SSD network on the image and post-processing the output using common algorithms (top-k filtering and Non-Maximum Suppression algorithm).

Here are two examples of successful detection outputs: [外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-CjkhuAAY-1597110071089)(pictures/ex1.png “SSD anchors”)] [外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-gQCnVYJn-1597110071091)(pictures/ex2.png “SSD anchors”)]

To run the notebook you first have to unzip the checkpoint files in ./checkpoint

and then start a jupyter notebook with

The current version only supports Pascal VOC datasets (2007 and 2012). In order to be used for training a SSD model, the former need to be converted to TF-Records using the tf_convert_data.py script:

Note the previous command generated a collection of TF-Records instead of a single file in order to ease shuffling during training.

The present TensorFlow implementation of SSD models have the following performances:

We are working hard at reproducing the same performance as the original Caffe implementation!

After downloading and extracting the previous checkpoints, the evaluation metrics should be reproducible by running the following command:

The evaluation script provides estimates on the recall-precision curve and compute the mAP metrics following the Pascal VOC 2007 and 2012 guidelines.

In addition, if one wants to experiment/test a different Caffe SSD checkpoint, the former can be converted to TensorFlow checkpoints as following:

The script train_ssd_network.py is in charged of training the network. Similarly to TF-Slim models, one can pass numerous options to the training process (dataset, optimiser, hyper-parameters, model, …). In particular, it is possible to provide a checkpoint file which can be use as starting point in order to fine-tune a network.

The easiest way to fine the SSD model is to use as pre-trained SSD network (VGG-300 or VGG-512). For instance, one can fine a model starting from the former as following:

Note that in addition to the training script flags, one may also want to experiment with data augmentation parameters (random cropping, resolution, …) in ssd_vgg_preprocessing.py or/and network parameters (feature layers, anchors boxes, …) in ssd_vgg_300/512.py

Furthermore, the training script can be combined with the evaluation routine in order to monitor the performance of saved checkpoints on a validation dataset. For that purpose, one can pass to training and validation scripts a GPU memory upper limit such that both can run in parallel on the same device. If some GPU memory is available for the evaluation script, the former can be run in parallel as follows:

One can also try to build a new SSD model based on standard architecture (VGG, ResNet, Inception, …) and set up on top of it the multibox layers (with specific anchors, ratios, …). For that purpose, you can fine-tune a network by only loading the weights of the original architecture, and initialize randomly the rest of network. For instance, in the case of the VGG-16 architecture, one can train a new model as following:

Hence, in the former command, the training script randomly initializes the weights belonging to the checkpoint_exclude_scopes and load from the checkpoint file vgg_16.ckpt the remaining part of the network. Note that we also specify with the trainable_scopes parameter to first only train the new SSD components and left the rest of VGG network unchanged. Once the network has converged to a good first result (~0.5 mAP for instance), you can fine-tuned the complete network as following:

A number of pre-trained weights of popular deep architectures can be found on TF-Slim models page.