标志检测和识别模型神经网络和代码

# 地标检测和侧面任务检测算法 ## 本地部署启动训练任务 ``` python3 mobile_ssd.py ``` ## 相关配置修改 ``` python train_parameters = { # 输入网络的图像尺寸 "input_size": [3, 300, 300], # 包含背景类的所有类别的个数 "class_dim": 6, "label_dict": {}, "image_count": -1, "log_feed_image": False, # 是否加载预训练模型 "pretrained": True, #预训练模型的路径 "pretrained_model_dir": "./pretrained-model", "continue_train": False, # 训练好的模型保存路径 "save_model_dir": "./ssd-model", "model_prefix": "mobilenet-ssd", # 数据所在根目录 "data_dir": "data/", # 预处理均值和方差 "mean_rgb": [127.5, 127.5, 127.5], # 训练标注文件相对于"data_dir"目录 "file_list": "road_train.txt", # 同理评测标注文件 "eval_list": "road_eval.txt" "mode": "train", "multi_data_reader_count": 5, # 训练轮数 "num_epochs": 120, # 训练batch "train_batch_size": 2, # 是否使用GPU "use_gpu": False, "apply_distort": True, "apply_expand": True, "apply_corp": True, "image_distort_strategy": { "expand_prob": 0.5, "expand_max_ratio": 4, "hue_prob": 0.5, "hue_delta": 18, "contrast_prob": 0.5, "contrast_delta": 0.5, "saturation_prob": 0.5, "saturation_delta": 0.5, "brightness_prob": 0.5, "brightness_delta": 0.125 }, "rsm_strategy": { "learning_rate": 0.001, "lr_epochs": [20, 40, 60, 80, 100], "lr_decay": [1, 0.5, 0.25, 0.1, 0.05, 0.01], }, "momentum_strategy": { "learning_rate": 0.1, "decay_steps": 2 ** 7, "decay_rate": 0.8 }, "early_stop": { "sample_frequency": 50, "successive_limit": 3, "min_loss": 1.28, "min_curr_map": 0.86 } } ``` ## 训练显卡配置在mobile_ssd.py ``` place = fluid.CUDAPlace(0) if train_parameters['use_gpu'] else fluid.CPUPlace() ``` ## 数据标志格式 见示例