重新训练MobilnetSSD v1 TensorFlow时结果不佳

Question

我只想为一个班级（人）从Model Zoo中重新训练ssd_mobilenet_v1_0.75_depth_coco。

我从COCO 2017集创建了TF记录。这给出了64115个训练示例和2693个验证示例。

我在Google Colab上设置了一个使用对象检测API的笔记本。训练有效，但我的成绩确实很差：

 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.001
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.003
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.002
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.005
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.049
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.001
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.131

这是在我添加到pipeline.config行fine_tune_checkpoint_type：'检测'后，如here所述，因为TF记录到缺少层。

配置中唯一的更改是：

train_config {
   batch_size: 24
   num_steps: 500
   fine_tune_checkpoint_type: 'detection'
}
eval_config {
   num_examples: 2693
}

当然添加路径。

我这样运行model_main.py

!python object_detection/model_main.py \
   --pipeline_config_path="pipeline.config" \
   --model_dir="training/" \
   --alsologtostderr \
   --num_train_steps=500

我将在更多步骤上运行此模型以处理整个数据集，但我认为12k示例足以获得更好的结果。知道这里发生了什么吗？

完整配置

model {
  ssd {
    num_classes: 1
    image_resizer {
      fixed_shape_resizer {
        height: 300
        width: 300
      }
    }
    feature_extractor {
      type: "ssd_mobilenet_v1"
      depth_multiplier: 0.75
      min_depth: 16
      conv_hyperparams {
        regularizer {
          l2_regularizer {
            weight: 3.99999989895e-05
          }
        }
        initializer {
          truncated_normal_initializer {
            mean: 0.0
            stddev: 0.0299999993294
          }
        }
        activation: RELU_6
        batch_norm {
          decay: 0.97000002861
          center: true
          scale: true
          epsilon: 0.0010000000475
          train: true
        }
      }
      override_base_feature_extractor_hyperparams: true
    }
    box_coder {
      faster_rcnn_box_coder {
        y_scale: 10.0
        x_scale: 10.0
        height_scale: 5.0
        width_scale: 5.0
      }
    }
    matcher {
      argmax_matcher {
        matched_threshold: 0.5
        unmatched_threshold: 0.5
        ignore_thresholds: false
        negatives_lower_than_unmatched: true
        force_match_for_each_row: true
        use_matmul_gather: true
      }
    }
    similarity_calculator {
      iou_similarity {
      }
    }
    box_predictor {
      convolutional_box_predictor {
        conv_hyperparams {
          regularizer {
            l2_regularizer {
              weight: 3.99999989895e-05
            }
          }
          initializer {
            random_normal_initializer {
              mean: 0.0
              stddev: 0.00999999977648
            }
          }
          activation: RELU_6
          batch_norm {
            decay: 0.97000002861
            center: true
            scale: true
            epsilon: 0.0010000000475
            train: true
          }
        }
        min_depth: 0
        max_depth: 0
        num_layers_before_predictor: 0
        use_dropout: false
        dropout_keep_probability: 0.800000011921
        kernel_size: 1
        box_code_size: 4
        apply_sigmoid_to_scores: false
        class_prediction_bias_init: -4.59999990463
      }
    }
    anchor_generator {
      ssd_anchor_generator {
        num_layers: 6
        min_scale: 0.20000000298
        max_scale: 0.949999988079
        aspect_ratios: 1.0
        aspect_ratios: 2.0
        aspect_ratios: 0.5
        aspect_ratios: 3.0
        aspect_ratios: 0.333299994469
      }
    }
    post_processing {
      batch_non_max_suppression {
        score_threshold: 0.300000011921
        iou_threshold: 0.600000023842
        max_detections_per_class: 100
        max_total_detections: 100
      }
      score_converter: SIGMOID
    }
    normalize_loss_by_num_matches: true
    loss {
      localization_loss {
        weighted_smooth_l1 {
          delta: 1.0
        }
      }
      classification_loss {
        weighted_sigmoid_focal {
          gamma: 2.0
          alpha: 0.75
        }
      }
      classification_weight: 1.0
      localization_weight: 1.0
    }
    encode_background_as_zeros: true
    normalize_loc_loss_by_codesize: true
    inplace_batchnorm_update: true
    freeze_batchnorm: false
  }
}
train_config {
  batch_size: 24
  data_augmentation_options {
    random_horizontal_flip {
    }
  }
  data_augmentation_options {
    ssd_random_crop {
    }
  }
  sync_replicas: true
  optimizer {
    momentum_optimizer {
      learning_rate {
        cosine_decay_learning_rate {
          learning_rate_base: 0.899999976158
          total_steps: 10000
          warmup_learning_rate: 0.300000011921
          warmup_steps: 300
        }
      }
      momentum_optimizer_value: 0.899999976158
    }
    use_moving_average: false
  }
  fine_tune_checkpoint: "/content/ssd_mobilenet_v1_0.75_depth_300x300_coco14_sync_2018_07_03/model.ckpt"
  fine_tune_checkpoint_type: 'detection'
  num_steps: 500
  startup_delay_steps: 0.0
  replicas_to_aggregate: 8
  max_number_of_boxes: 100
  unpad_groundtruth_tensors: false
}
train_input_reader {
  label_map_path: "/content/classes.pbtxt"
  tf_record_input_reader {
    input_path: "/content/gdrive/My Drive/coco_train.record"
  }
}
eval_config {
  num_examples: 2693
  metrics_set: "coco_detection_metrics"
  use_moving_averages: false
}
eval_input_reader {
  label_map_path: "/content/classes.pbtxt"
  shuffle: false
  num_readers: 1
  tf_record_input_reader {
    input_path: "/content/gdrive/My Drive/coco_val.record"
  }
}

图

编辑在27k / 53k步骤之后

 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.005
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.020
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.001
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.000
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.001
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.012
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.016
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.048
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.048
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.000
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.130

基本上它在评估图像上放了一个具有51％概率的盒子（这里只是一个例子：]

Answer 1

我正在训练一个单一类别的对象检测模型（以检测火）。我已经使用SSD mobileNet V2训练了模型。我已经在Google colab上对其进行了培训。我正在寻找一种计算平均精度的方法。有人可以分享方法/代码吗？我真的很困惑，因为它是单个类检测。我是机器学习的新手。谢谢