{
"info": {
"author": "Gluon CV Toolkit Contributors",
"author_email": "",
"bugtrack_url": null,
"classifiers": [],
"description": "GluonCV-Torch\n=============\n\nLoad `GluonCV `__ Models in PyTorch. Simply\n``import gluoncvth`` to getting better pretrained model than\n``torchvision``:\n\n.. code:: python\n\n import gluoncvth as gcv\n model = gcv.models.resnet50(pretrained=True)\n\n**Installation**:\n\n.. code:: bash\n\n pip install gluoncv-torch\n\nAvailable Models\n----------------\n\nImageNet\n~~~~~~~~\n\nImageNet models single-crop error rates, comparing to the\n``torchvision`` models:\n\n+---------------------------+---------------+---------------+---------------+---------------+\n| | torchvision | | gluoncvth | |\n+===========================+===============+===============+===============+===============+\n| Model | Top-1 error | Top-5 error | Top-1 error | Top-5 error |\n+---------------------------+---------------+---------------+---------------+---------------+\n| `ResNet18 <#resnet>`__ | 30.24 | 10.92 | 29.06 | 10.17 |\n+---------------------------+---------------+---------------+---------------+---------------+\n| `ResNet34 <#resnet>`__ | 26.70 | 8.58 | 25.35 | 7.92 |\n+---------------------------+---------------+---------------+---------------+---------------+\n| `ResNet50 <#resnet>`__ | 23.85 | 7.13 | 22.33 | 6.18 |\n+---------------------------+---------------+---------------+---------------+---------------+\n| `ResNet101 <#resnet>`__ | 22.63 | 6.44 | 20.80 | 5.39 |\n+---------------------------+---------------+---------------+---------------+---------------+\n| `ResNet152 <#resnet>`__ | 21.69 | 5.94 | 20.56 | 5.39 |\n+---------------------------+---------------+---------------+---------------+---------------+\n| Inception v3 | 22.55 | 6.44 | 21.33 | 5.61 |\n+---------------------------+---------------+---------------+---------------+---------------+\n\nMore models available at `GluonCV Image Classification\nModelZoo `__\n\nSemantic Segmentation\n~~~~~~~~~~~~~~~~~~~~~\n\nResults on Pascal VOC dataset:\n\n+------------------------------+----------------+--------+\n| Model | Base Network | mIoU |\n+==============================+================+========+\n| `FCN <#fcn>`__ | ResNet101 | 83.6 |\n+------------------------------+----------------+--------+\n| `PSPNet <#pspnet>`__ | ResNet101 | 85.1 |\n+------------------------------+----------------+--------+\n| `DeepLabV3 <#deeplabv3>`__ | ResNet101 | 86.2 |\n+------------------------------+----------------+--------+\n\nResults on ADE20K dataset:\n\n+------------------------------+----------------+----------+--------+\n| Model | Base Network | PixAcc | mIoU |\n+==============================+================+==========+========+\n| `FCN <#fcn>`__ | ResNet101 | 80.6 | 41.6 |\n+------------------------------+----------------+----------+--------+\n| `PSPNet <#pspnet>`__ | ResNet101 | 80.8 | 42.9 |\n+------------------------------+----------------+----------+--------+\n| `DeepLabV3 <#deeplabv3>`__ | ResNet101 | 81.1 | 44.1 |\n+------------------------------+----------------+----------+--------+\n\n**Quick Demo**\n\n.. code:: python\n\n import torch\n import gluoncvth\n\n # Get the model\n model = gluoncvth.models.get_deeplab_resnet101_ade(pretrained=True)\n model.eval()\n\n # Prepare the image\n url = 'https://github.com/zhanghang1989/image-data/blob/master/encoding/' + \\\n 'segmentation/ade20k/ADE_val_00001142.jpg?raw=true'\n filename = 'example.jpg'\n img = gluoncvth.utils.load_image(\n gluoncvth.utils.download(url, filename)).unsqueeze(0)\n\n # Make prediction\n output = model.evaluate(img)\n predict = torch.max(output, 1)[1].cpu().numpy() + 1\n\n # Get color pallete for visualization\n mask = gluoncvth.utils.get_mask_pallete(predict, 'ade20k')\n mask.save('output.png')\n\n.. figure:: ./image/demo_deeplab_ade.png\n :alt: \n\nMore models available at `GluonCV Semantic Segmentation\nModelZoo `__\n\nAPI Reference\n-------------\n\nResNet\n~~~~~~\n\n- ``gluoncvth.models.resnet18(pretrained=True)``\n- ``gluoncvth.models.resnet34(pretrained=True)``\n- ``gluoncvth.models.resnet50(pretrained=True)``\n- ``gluoncvth.models.resnet101(pretrained=True)``\n- ``gluoncvth.models.resnet152(pretrained=True)``\n\nFCN\n~~~\n\n- ``gluoncvth.models.get_fcn_resnet101_voc(pretrained=True)``\n- ``gluoncvth.models.get_fcn_resnet101_ade(pretrained=True)``\n\nPSPNet\n~~~~~~\n\n- ``gluoncvth.models.get_psp_resnet101_voc(pretrained=True)``\n- ``gluoncvth.models.get_psp_resnet101_ade(pretrained=True)``\n\nDeepLabV3\n~~~~~~~~~\n\n- ``gluoncvth.models.get_deeplab_resnet101_voc(pretrained=True)``\n- ``gluoncvth.models.get_deeplab_resnet101_ade(pretrained=True)``\n\nWhy `GluonCV `__?\n---------------------------------------------\n\n**1. State-of-the-art Implementations**\n\n**2. Pretrained Models and Tutorials**\n\n**3. Community Support**\n\nWe expect this PyTorch inference API for GluonCV models will be\nbeneficial to the entire computer vision comunity.",
"description_content_type": "",
"docs_url": null,
"download_url": "",
"downloads": {
"last_day": -1,
"last_month": -1,
"last_week": -1
},
"home_page": "https://github.com/dmlc/gluon-cv",
"keywords": "",
"license": "Apache-2.0",
"maintainer": "",
"maintainer_email": "",
"name": "gluoncv-torch",
"package_url": "https://pypi.org/project/gluoncv-torch/",
"platform": "",
"project_url": "https://pypi.org/project/gluoncv-torch/",
"project_urls": {
"Homepage": "https://github.com/dmlc/gluon-cv"
},
"release_url": "https://pypi.org/project/gluoncv-torch/0.0.3/",
"requires_dist": null,
"requires_python": "",
"summary": "MXNet Gluon CV Toolkit",
"version": "0.0.3"
},
"last_serial": 4384376,
"releases": {
"0.0.0": [
{
"comment_text": "",
"digests": {
"md5": "c202c7b2f1287b698cbcc6f567e2b41e",
"sha256": "a9b85f645e857306aaec8b9444dfe143ed6436dbdcde2edc783be404099b24ca"
},
"downloads": -1,
"filename": "gluoncv-torch-0.0.0.tar.gz",
"has_sig": false,
"md5_digest": "c202c7b2f1287b698cbcc6f567e2b41e",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 1499,
"upload_time": "2018-10-09T22:35:57",
"url": "https://files.pythonhosted.org/packages/dd/bd/c5fc328bb4d608399993025a0aeffdd1c3f67bddd7701838d267aa54be78/gluoncv-torch-0.0.0.tar.gz"
}
],
"0.0.2": [
{
"comment_text": "",
"digests": {
"md5": "0c923d9bfe00823c8ef48dbf5de3f8c3",
"sha256": "314aff16d1a9fe6b8e9f0d54ce9f9ebf02ffbd90eedabd9615f68f402f9fee7b"
},
"downloads": -1,
"filename": "gluoncv-torch-0.0.2.tar.gz",
"has_sig": false,
"md5_digest": "0c923d9bfe00823c8ef48dbf5de3f8c3",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 15875,
"upload_time": "2018-10-12T23:21:42",
"url": "https://files.pythonhosted.org/packages/93/17/ee76d47c8e1256ac4c8c5101812a8eb55c0ee44a52fa03d9da88b3ddc94e/gluoncv-torch-0.0.2.tar.gz"
}
],
"0.0.3": [
{
"comment_text": "",
"digests": {
"md5": "6986546ecc3f31e79729ee4b32886430",
"sha256": "648182fac244703c02b5853b9f34f5d8afc55fb600c3090a55006e7e2f2c09c5"
},
"downloads": -1,
"filename": "gluoncv-torch-0.0.3.tar.gz",
"has_sig": false,
"md5_digest": "6986546ecc3f31e79729ee4b32886430",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 15821,
"upload_time": "2018-10-16T23:28:56",
"url": "https://files.pythonhosted.org/packages/04/1c/e919d53a5b4d12fc8aeaead5097889eab57a05c92b800d03c0bdded4f220/gluoncv-torch-0.0.3.tar.gz"
}
]
},
"urls": [
{
"comment_text": "",
"digests": {
"md5": "6986546ecc3f31e79729ee4b32886430",
"sha256": "648182fac244703c02b5853b9f34f5d8afc55fb600c3090a55006e7e2f2c09c5"
},
"downloads": -1,
"filename": "gluoncv-torch-0.0.3.tar.gz",
"has_sig": false,
"md5_digest": "6986546ecc3f31e79729ee4b32886430",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 15821,
"upload_time": "2018-10-16T23:28:56",
"url": "https://files.pythonhosted.org/packages/04/1c/e919d53a5b4d12fc8aeaead5097889eab57a05c92b800d03c0bdded4f220/gluoncv-torch-0.0.3.tar.gz"
}
]
}