Waste Allocation & Recognition Project
Images 0.jpg to 555.jpg are part of wade-dataset, all the other images are from google street view and
annotated by me, because of the little time I had there should be some mistakes, feel free to email me
if you find any!
This model was trained using the pretrained model: ssd_mobilenet_v1_fpn_640x640_coco17_tpu-8,
to run this model download the tensorflow-object-detection api.
Load the model using the script object_detection_tutorial.ipynb in models/research/object_detection/colab_tutorials/
replace PATH_TO_LABELS = 'models/research/object_detection/data/mscoco_label_map.pbtxt'
with the directory of the label_map.pbtxt that I provide in the repo.
replace detection_model = load_model(model_name) with detection_model = tf.saved_model.load('__path to pile_model__')
you are ready to go!
Contact: www.maestrostudio.org