Question

系统信息

您正在使用的模型的顶级目录是什么： object_detection / ssd_inception_v2以及slim / InceptionV2
我是否编写过自定义代码（与使用TensorFlow中提供的库存示例脚本相反）：否
OS平台和发行版（例如，Linux Ubuntu 16.04）： Ubuntu 16.04
从（源代码或二进制代码）安装的TensorFlow：二进制文件
TensorFlow版本（使用下面的命令）： 1.2.1
Bazel版本（如果从源代码编译）：否
CUDA / cuDNN版本： cuda 8.0
GPU型号和内存： Quadro M6000 24GB
重现的确切命令：

我在（交通标志）分类数据集上训练了一个inceptionV2（来自slim）。然后，我想使用生成的检查点作为对象检测API的基础。如果我理解正确，应该可以使用/ object_detection / samples / configs / ssd_inception_v2_pets.config配置文件的版本。但是，当我运行对象检测时

python object_detection / train.py --logtostderr --pipeline_config_path = / home / tobi / tensorflow / trafficsigns / models / model / ssd_inception_v2_trafficsigns.config -train_dir = / home / tobi / tensorflow / trafficsigns / models / model /列车

我收到以下错误

Traceback (most recent call last):
  File "object_detection/train.py", line 198, in <module>
    tf.app.run()
  File "/home/tobi/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/platform/app.py", line 48, in run
    _sys.exit(main(_sys.argv[:1] + flags_passthrough))
  File "object_detection/train.py", line 194, in main
    worker_job_name, is_chief, FLAGS.train_dir)
  File "/home/tobi/tensorflow/lib/python2.7/site-packages/tensorflow/models/object_detection/trainer.py", line 191, in train
    clones = model_deploy.create_clones(deploy_config, model_fn, [input_queue])
  File "/home/tobi/tensorflow/local/lib/python2.7/site-packages/tensorflow/models/slim/deployment/model_deploy.py", line 193, in create_clones
    outputs = model_fn(*args, **kwargs)
  File "/home/tobi/tensorflow/lib/python2.7/site-packages/tensorflow/models/object_detection/trainer.py", line 132, in _create_losses
    losses_dict = detection_model.loss(prediction_dict)
  File "/home/tobi/tensorflow/lib/python2.7/site-packages/tensorflow/models/object_detection/meta_architectures/ssd_meta_arch.py", line 431, in loss
    location_losses, cls_losses, prediction_dict, match_list)
  File "/home/tobi/tensorflow/lib/python2.7/site-packages/tensorflow/models/object_detection/meta_architectures/ssd_meta_arch.py", line 551, in _apply_hard_mining
    [0, 0, 1], class_pred_shape), class_pred_shape)
  File "/home/tobi/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/ops/array_ops.py", line 547, in slice
    return gen_array_ops._slice(input_, begin, size, name=name)
  File "/home/tobi/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/ops/gen_array_ops.py", line 2896, in _slice
    name=name)
  File "/home/tobi/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/framework/op_def_library.py", line 499, in apply_op
    repr(values), type(values).__name__))
TypeError: Expected int32 passed to parameter 'size' of op 'Slice', got [-1, None, 43] of type 'list' instead.

在我看来，问题是锚框没有被初始化，导致'None'值让tf.Slice操作崩溃。

我使用的配置文件

 # SSD with Inception v2 configured for Oxford-IIIT Pets Dataset.
    # Users should configure the fine_tune_checkpoint field in the train config as
    # well as the label_map_path and input_path fields in the train_input_reader and
    # eval_input_reader. Search for "PATH_TO_BE_CONFIGURED" to find the fields that
    # should be configured.

    model {
      ssd {
        num_classes: 43
        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
          }
        }
        similarity_calculator {
          iou_similarity {
          }
        }
        anchor_generator {
          ssd_anchor_generator {
            num_layers: 6
            min_scale: 0.2
            max_scale: 0.95
            aspect_ratios: 1.0
            aspect_ratios: 2.0
            aspect_ratios: 0.5
            aspect_ratios: 3.0
            aspect_ratios: 0.3333
            reduce_boxes_in_lowest_layer: true
          }
        }
        image_resizer {
           keep_aspect_ratio_resizer {
            min_dimension: 400
            max_dimension: 690
          }
        }
        box_predictor {
          convolutional_box_predictor {
            min_depth: 0
            max_depth: 0
            num_layers_before_predictor: 0
            use_dropout: false
            dropout_keep_probability: 0.8
            kernel_size: 3
            box_code_size: 4
            apply_sigmoid_to_scores: false
            conv_hyperparams {
              activation: RELU_6,
              regularizer {
                l2_regularizer {
                  weight: 0.00004
                }
              }
              initializer {
                truncated_normal_initializer {
                  stddev: 0.03
                  mean: 0.0
                }
              }
            }
          }
        }
        feature_extractor {
          type: 'ssd_inception_v2'
          min_depth: 16
          depth_multiplier: 1.0
          conv_hyperparams {
            activation: RELU_6,
            regularizer {
              l2_regularizer {
                weight: 0.00004
              }
            }
            initializer {
              truncated_normal_initializer {
                stddev: 0.03
                mean: 0.0
              }
            }
            batch_norm {
              train: true,
              scale: true,
              center: true,
              decay: 0.9997,
              epsilon: 0.001,
            }
          }
        }
        loss {
          classification_loss {
            weighted_sigmoid {
              anchorwise_output: true
            }
          }
          localization_loss {
            weighted_smooth_l1 {
              anchorwise_output: true
            }
          }
          hard_example_miner {
            num_hard_examples: 3000
            iou_threshold: 0.99
            loss_type: CLASSIFICATION
            max_negatives_per_positive: 3
            min_negatives_per_image: 0
          }
          classification_weight: 1.0
          localization_weight: 1.0
        }
        normalize_loss_by_num_matches: true
        post_processing {
          batch_non_max_suppression {
            score_threshold: 1e-8
            iou_threshold: 0.6
            max_detections_per_class: 100
            max_total_detections: 100
          }
          score_converter: SIGMOID
        }
      }
    }

    train_config: {
      batch_size: 24
      optimizer {
        rms_prop_optimizer: {
          learning_rate: {
            exponential_decay_learning_rate {
              initial_learning_rate: 0.004
              decay_steps: 800720
              decay_factor: 0.95
            }
          }
          momentum_optimizer_value: 0.9
          decay: 0.9
          epsilon: 1.0
        }
      }
      fine_tune_checkpoint: "/home/tobi/tensorflow/trafficsigns/models/pretrained/inception_v2_GTS/model.ckpt"
      from_detection_checkpoint: false
      # Note: The below line limits the training process to 200K steps, which we
      # empirically found to be sufficient enough to train the pets dataset. This
      # effectively bypasses the learning rate schedule (the learning rate will
      # never decay). Remove the below line to train indefinitely.
      #num_steps: 200000
      data_augmentation_options {
        random_horizontal_flip {
        }
      }
      data_augmentation_options {
        ssd_random_crop {
        }
      }
    }

    train_input_reader: {
      tf_record_input_reader {
        input_path: "/home/tobi/tensorflow/trafficsigns/data/train.record"
      }
      label_map_path: "/home/tobi/tensorflow/trafficsigns/data/ts_label_map.pbtxt"
    }

    eval_config: {
      num_examples: 2000
      num_visualizations: 25
      # Note: The below line limits the evaluation process to 10 evaluations.
      # Remove the below line to evaluate indefinitely.
      max_evals: 10
    }

    eval_input_reader: {
      tf_record_input_reader {
        input_path: "/home/tobi/tensorflow/trafficsigns/data/test.record"
      }
      label_map_path: "/home/tobi/tensorflow/trafficsigns/data/ts_label_map.pbtxt"
      shuffle: false
      num_readers: 1
    }

任何想法出了什么问题？提前谢谢。

修改

按照建议（再次）将 keep_aspect_ratio_resizer 更改为配置文件中的 fixed_shape_resizer 后，培训过程开始。我不知怎么得到两次记录消息，这不会让我感到烦恼。我担心的是以下警告，表明检查点不包含预期变量。无论如何，培训仍然有效，我还没有检查结果，但我不确定是否应该忽略这些。

WARNING:root:Variable [InceptionV2/Conv2d_1a_7x7/BatchNorm/beta] not available in checkpoint
WARNING:root:Variable [InceptionV2/Conv2d_1a_7x7/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Conv2d_1a_7x7/BatchNorm/moving_mean] not available in checkpoint
WARNING:root:Variable [InceptionV2/Conv2d_1a_7x7/BatchNorm/moving_variance] not available in checkpoint
WARNING:root:Variable [InceptionV2/Conv2d_2b_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Conv2d_2c_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_3b/Branch_0/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_3b/Branch_1/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_3b/Branch_1/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_3b/Branch_2/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_3b/Branch_2/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_3b/Branch_2/Conv2d_0c_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_3b/Branch_3/Conv2d_0b_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_3c/Branch_0/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_3c/Branch_1/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_3c/Branch_1/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_3c/Branch_2/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_3c/Branch_2/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_3c/Branch_2/Conv2d_0c_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_3c/Branch_3/Conv2d_0b_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4a/Branch_0/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4a/Branch_0/Conv2d_1a_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4a/Branch_1/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4a/Branch_1/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4a/Branch_1/Conv2d_1a_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4b/Branch_0/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4b/Branch_1/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4b/Branch_1/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4b/Branch_2/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4b/Branch_2/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4b/Branch_2/Conv2d_0c_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4b/Branch_3/Conv2d_0b_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4c/Branch_0/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4c/Branch_1/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4c/Branch_1/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4c/Branch_2/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4c/Branch_2/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4c/Branch_2/Conv2d_0c_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4c/Branch_3/Conv2d_0b_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4d/Branch_0/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4d/Branch_1/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4d/Branch_1/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4d/Branch_2/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4d/Branch_2/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4d/Branch_2/Conv2d_0c_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4d/Branch_3/Conv2d_0b_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4e/Branch_0/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4e/Branch_1/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4e/Branch_1/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4e/Branch_2/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4e/Branch_2/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4e/Branch_2/Conv2d_0c_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_4e/Branch_3/Conv2d_0b_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5a/Branch_0/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5a/Branch_0/Conv2d_1a_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5a/Branch_1/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5a/Branch_1/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5a/Branch_1/Conv2d_1a_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5b/Branch_0/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5b/Branch_1/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5b/Branch_1/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5b/Branch_2/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5b/Branch_2/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5b/Branch_2/Conv2d_0c_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5b/Branch_3/Conv2d_0b_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c/Branch_0/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c/Branch_1/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c/Branch_1/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c/Branch_2/Conv2d_0a_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c/Branch_2/Conv2d_0b_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c/Branch_2/Conv2d_0c_3x3/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c/Branch_3/Conv2d_0b_1x1/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_2_1x1_256/BatchNorm/beta] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_2_1x1_256/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_2_1x1_256/BatchNorm/moving_mean] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_2_1x1_256/BatchNorm/moving_variance] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_2_1x1_256/weights] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_3_1x1_128/BatchNorm/beta] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_3_1x1_128/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_3_1x1_128/BatchNorm/moving_mean] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_3_1x1_128/BatchNorm/moving_variance] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_3_1x1_128/weights] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_4_1x1_128/BatchNorm/beta] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_4_1x1_128/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_4_1x1_128/BatchNorm/moving_mean] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_4_1x1_128/BatchNorm/moving_variance] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_4_1x1_128/weights] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_5_1x1_64/BatchNorm/beta] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_5_1x1_64/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_5_1x1_64/BatchNorm/moving_mean] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_5_1x1_64/BatchNorm/moving_variance] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_1_Conv2d_5_1x1_64/weights] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_2_3x3_s2_512/BatchNorm/beta] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_2_3x3_s2_512/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_2_3x3_s2_512/BatchNorm/moving_mean] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_2_3x3_s2_512/BatchNorm/moving_variance] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_2_3x3_s2_512/weights] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_3_3x3_s2_256/BatchNorm/beta] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_3_3x3_s2_256/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_3_3x3_s2_256/BatchNorm/moving_mean] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_3_3x3_s2_256/BatchNorm/moving_variance] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_3_3x3_s2_256/weights] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_4_3x3_s2_256/BatchNorm/beta] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_4_3x3_s2_256/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_4_3x3_s2_256/BatchNorm/moving_mean] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_4_3x3_s2_256/BatchNorm/moving_variance] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_4_3x3_s2_256/weights] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_5_3x3_s2_128/BatchNorm/beta] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_5_3x3_s2_128/BatchNorm/gamma] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_5_3x3_s2_128/BatchNorm/moving_mean] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_5_3x3_s2_128/BatchNorm/moving_variance] not available in checkpoint
WARNING:root:Variable [InceptionV2/Mixed_5c_2_Conv2d_5_3x3_s2_128/weights] not available in checkpoint

Answer 1

我发现上述配置至少存在一个问题 - 这就是keep_aspect_ratio_resizer与大于1的批量大小不兼容。（我知道这还没有在任何地方记录）。问题在于，在调整每个图像的大小后，由于保留了纵横比，所有图像都具有不同的形状，因此我们无法将它们堆叠成批处理。

我建议使用固定形状，如300x300。

Answer 2

在您的配置文件中，将火车配置中的fine_tune_checkpoint_type设置为"classification"并将批处理规范中的scale和center设置为false，如下所示：< / p>

train_config {
    fine_tune_checkpoint: "model.ckpt",
    fine_tune_checkpoint_type: "classification"
    from_detection_checkpoint: false
}


batch_norm {
    train: true,
    scale: false,
    center: false,
    decay: 0.9997,
    epsilon: 0.001,
}

Tensorflow：使用分类检查点进行对象检测

2 个答案: