Project: HereIsWally (GitHub Link)

HereIsWally-master
- images
  - .DS_Store
- LICENSE
- Pipfile
- find_wally_pretty.py
- annotations
  - annotations.csv
- README.md
- object_detection
  - BUILD
  - matchers
    - BUILD
    - bipartite_matcher_test.py
    - argmax_matcher.py
    - __init__.py
    - argmax_matcher_test.py
    - bipartite_matcher.py
  - object_detection_tutorial.ipynb
  - create_pascal_tf_record_test.py
  - evaluator.py
  - samples
    - cloud
      - cloud.yml
    - configs
      - faster_rcnn_resnet101_voc07.config
      - faster_rcnn_resnet101_pets.config
      - faster_rcnn_resnet50_pets.config
      - ssd_inception_v2_pets.config
      - rfcn_resnet101_pets.config
      - faster_rcnn_resnet152_pets.config
      - ssd_mobilenet_v1_pets.config
      - faster_rcnn_inception_resnet_v2_atrous_pets.config
  - ssd_mobilenet_v1_coco_11_06_2017
  - utils
    - BUILD
    - label_map_util_test.py
    - test_utils.py
    - np_box_list_test.py
    - __pycache__
      - visualization_utils.cpython-36.pyc
      - __init__.cpython-36.pyc
      - label_map_util.cpython-36.pyc
    - ops.py
    - metrics_test.py
    - metrics.py
    - category_util.py
    - visualization_utils.py
    - shape_utils.py
    - label_map_util.py
    - dataset_util.py
    - variables_helper.py
    - test_utils_test.py
    - np_box_list_ops.py
    - category_util_test.py
    - learning_schedules.py
    - static_shape.py
    - np_box_ops.py
    - dataset_util_test.py
    - variables_helper_test.py
    - visualization_utils_test.py
    - learning_schedules_test.py
    - __init__.py
    - static_shape_test.py
    - shape_utils_test.py
    - ops_test.py
    - per_image_evaluation.py
    - np_box_list_ops_test.py
    - np_box_ops_test.py
    - np_box_list.py
    - object_detection_evaluation_test.py
    - per_image_evaluation_test.py
    - object_detection_evaluation.py
  - test_images
    - image_info.txt
  - eval.py
  - g3doc
    - running_on_cloud.md
    - running_pets.md
    - configuring_jobs.md
    - img
    - defining_your_own_model.md
    - running_locally.md
    - running_notebook.md
    - installation.md
    - detection_model_zoo.md
    - preparing_inputs.md
    - exporting_models.md
  - __pycache__
    - __init__.cpython-36.pyc
  - builders
    - anchor_generator_builder.py
    - BUILD
    - optimizer_builder_test.py
    - image_resizer_builder_test.py
    - anchor_generator_builder_test.py
    - post_processing_builder_test.py
    - region_similarity_calculator_builder_test.py
    - preprocessor_builder_test.py
    - box_coder_builder.py
    - optimizer_builder.py
    - losses_builder_test.py
    - hyperparams_builder.py
    - matcher_builder.py
    - image_resizer_builder.py
    - post_processing_builder.py
    - box_predictor_builder.py
    - box_predictor_builder_test.py
    - input_reader_builder_test.py
    - hyperparams_builder_test.py
    - model_builder_test.py
    - losses_builder.py
    - __init__.py
    - input_reader_builder.py
    - preprocessor_builder.py
    - matcher_builder_test.py
    - box_coder_builder_test.py
    - region_similarity_calculator_builder.py
    - model_builder.py
  - create_pet_tf_record.py
  - protos
    - bipartite_matcher.proto
    - matcher_pb2.py
    - mean_stddev_box_coder_pb2.py
    - BUILD
    - faster_rcnn_box_coder.proto
    - train.proto
    - ssd_pb2.py
    - model.proto
    - faster_rcnn.proto
    - eval.proto
    - string_int_label_map.proto
    - preprocessor.proto
    - region_similarity_calculator_pb2.py
    - hyperparams.proto
    - train_pb2.py
    - box_predictor_pb2.py
    - region_similarity_calculator.proto
    - image_resizer.proto
    - hyperparams_pb2.py
    - __pycache__
      - __init__.cpython-36.pyc
      - string_int_label_map_pb2.cpython-36.pyc
    - input_reader_pb2.py
    - ssd_anchor_generator.proto
    - model_pb2.py
    - argmax_matcher_pb2.py
    - input_reader.proto
    - faster_rcnn_box_coder_pb2.py
    - grid_anchor_generator_pb2.py
    - post_processing_pb2.py
    - pipeline.proto
    - string_int_label_map_pb2.py
    - square_box_coder.proto
    - optimizer_pb2.py
    - eval_pb2.py
    - optimizer.proto
    - image_resizer_pb2.py
    - argmax_matcher.proto
    - losses.proto
    - box_predictor.proto
    - __init__.py
    - pipeline_pb2.py
    - square_box_coder_pb2.py
    - matcher.proto
    - losses_pb2.py
    - post_processing.proto
    - ssd_anchor_generator_pb2.py
    - ssd.proto
    - anchor_generator_pb2.py
    - box_coder.proto
    - grid_anchor_generator.proto
    - preprocessor_pb2.py
    - faster_rcnn_pb2.py
    - box_coder_pb2.py
    - anchor_generator.proto
    - bipartite_matcher_pb2.py
    - mean_stddev_box_coder.proto
  - models
    - faster_rcnn_inception_resnet_v2_feature_extractor_test.py
    - feature_map_generators.py
    - BUILD
    - ssd_feature_extractor_test.py
    - faster_rcnn_resnet_v1_feature_extractor_test.py
    - ssd_inception_v2_feature_extractor_test.py
    - ssd_inception_v2_feature_extractor.py
    - faster_rcnn_resnet_v1_feature_extractor.py
    - feature_map_generators_test.py
    - ssd_mobilenet_v1_feature_extractor_test.py
    - __init__.py
    - ssd_mobilenet_v1_feature_extractor.py
    - faster_rcnn_inception_resnet_v2_feature_extractor.py
  - CONTRIBUTING.md
  - anchor_generators
    - BUILD
    - multiple_grid_anchor_generator.py
    - multiple_grid_anchor_generator_test.py
    - __init__.py
    - grid_anchor_generator_test.py
    - grid_anchor_generator.py
  - exporter.py
  - export_inference_graph.py
  - data
    - pascal_label_map.pbtxt
    - pet_label_map.pbtxt
    - mscoco_label_map.pbtxt
  - meta_architectures
    - BUILD
    - faster_rcnn_meta_arch.py
    - faster_rcnn_meta_arch_test_lib.py
    - ssd_meta_arch.py
    - rfcn_meta_arch.py
    - faster_rcnn_meta_arch_test.py
    - __init__.py
    - rfcn_meta_arch_test.py
    - ssd_meta_arch_test.py
  - train.py
  - __init__.py
  - eval_util.py
  - README.md
  - core
    - BUILD
    - box_predictor_test.py
    - balanced_positive_negative_sampler.py
    - prefetcher.py
    - region_similarity_calculator.py
    - batcher.py
    - box_list_test.py
    - post_processing_test.py
    - matcher_test.py
    - minibatch_sampler_test.py
    - target_assigner.py
    - preprocessor.py
    - minibatch_sampler.py
    - data_decoder.py
    - keypoint_ops.py
    - keypoint_ops_test.py
    - losses.py
    - box_predictor.py
    - post_processing.py
    - box_list_ops.py
    - box_list.py
    - balanced_positive_negative_sampler_test.py
    - box_list_ops_test.py
    - model.py
    - matcher.py
    - standard_fields.py
    - region_similarity_calculator_test.py
    - box_coder.py
    - __init__.py
    - box_coder_test.py
    - preprocessor_test.py
    - target_assigner_test.py
    - batcher_test.py
    - anchor_generator.py
    - prefetcher_test.py
    - losses_test.py
  - trainer_test.py
  - trainer.py
  - create_pascal_tf_record.py
  - box_coders
    - BUILD
    - keypoint_box_coder.py
    - square_box_coder_test.py
    - faster_rcnn_box_coder_test.py
    - mean_stddev_box_coder.py
    - mean_stddev_box_coder_test.py
    - keypoint_box_coder_test.py
    - faster_rcnn_box_coder.py
    - __init__.py
    - square_box_coder.py
  - data_decoders
    - BUILD
    - tf_example_decoder_test.py
    - __init__.py
    - tf_example_decoder.py
  - exporter_test.py
- find_wally.py
- docs
- eval_images
- trained_model
  - saved_model
  - model.ckpt.index
  - labels.txt
  - example.config
  - checkpoint

HereIsWally

Deep learning project that solves Where's Wally puzzles by finding the exact position of Wally in an image

alt text

HereIsWally is a Tensorflow project that includes a model for solving Where's Wally puzzles. It uses Faster RCNN Inception v2 model initially trained on COCO dataset and retrained for finding Wally using transfer learning with Tensorflow Object Detection API.

Getting ready

Install Python 3.7
Install Pipenv
Run pipenv install to install the dependencies
Run pipenv shell to activate the Pipenv environment

Usage

python find_wally_pretty.py images/1.jpg

python find_wally.py images/1.jpg

A window displaying a Wally puzzle outlining Wally should show up.

Retraining the model

For instructions about how to retrain the model follow the instructions in this blog post.

HereIsWally

Getting ready

Usage

Retraining the model

Sources

Related projects

Copyright