- شرح :
ImageNet-PI یک نسخه تغییر برچسبگذاری شده از مجموعه داده استاندارد ILSVRC2012 ImageNet است که در آن برچسبها توسط مجموعهای از ۱۶ شبکه عصبی عمیق با معماریهای مختلف که از قبل روی ILSVRC2012 استاندارد آموزش دیدهاند، ارائه میشوند. به طور خاص، مدلهای از پیش آموزش دیده از tf.keras.applications دانلود میشوند.
علاوه بر برچسبهای جدید، ImageNet-PI همچنین متادادههایی در مورد فرآیند حاشیهنویسی به شکل محرمانگی مدلها روی برچسبهایشان و اطلاعات اضافی در مورد هر مدل ارائه میدهد.
برای اطلاعات بیشتر به ImageNet-PI مراجعه کنید.
صفحه اصلی : https://github.com/google-research-datasets/imagenet_pi/
کد منبع :
tfds.datasets.imagenet_pi.Builderنسخهها :
-
1.0.0(پیشفرض): انتشار اولیه.
-
حجم دانلود :
Unknown sizeاندازه مجموعه داده :
Unknown sizeدستورالعمل دانلود دستی : این مجموعه داده از شما میخواهد که دادههای منبع را به صورت دستی در
download_config.manual_dirدانلود کنید (پیشفرض روی~/tensorflow_datasets/downloads/manual/):
manual_dir باید شامل دو فایل باشد: ILSVRC2012_img_train.tar و ILSVRC2012_img_val.tar. برای دریافت لینک دانلود مجموعه دادهها، باید در http://www.image-net.org/download-images ثبتنام کنید.ذخیره خودکار ( مستندات ): ناشناخته
تقسیمات :
| تقسیم | مثالها |
|---|
- ساختار ویژگی :
FeaturesDict({
'annotator_confidences': Tensor(shape=(16,), dtype=float32),
'annotator_labels': Tensor(shape=(16,), dtype=int64),
'clean_label': ClassLabel(shape=(), dtype=int64, num_classes=1000),
'file_name': Text(shape=(), dtype=string),
'image': Image(shape=(None, None, 3), dtype=uint8),
})
- مستندات ویژگیها :
| ویژگی | کلاس | شکل | نوع D | توضیحات |
|---|---|---|---|---|
| ویژگیها | ||||
| annotator_confidences | تانسور | (16،) | شناور32 | |
| برچسبهای حاشیهنویس | تانسور | (16،) | int64 | |
| clean_label | برچسب کلاس | int64 | ||
| نام_فایل | متن | رشته | ||
| تصویر | تصویر | (هیچ، هیچ، ۳) | uint8 |
کلیدهای تحت نظارت (به سند
as_supervisedمراجعه کنید):('image', 'annotator_labels')شکل ( tfds.show_examples ): پشتیبانی نمیشود.
مثالها ( tfds.as_dataframe ): موجود نیست.
استناد :
@inproceedings{tram,
author = {Mark Collier and
Rodolphe Jenatton and
Effrosyni Kokiopoulou and
Jesse Berent},
editor = {Kamalika Chaudhuri and
Stefanie Jegelka and
Le Song and
Csaba Szepesv{\'{a} }ri and
Gang Niu and
Sivan Sabato},
title = {Transfer and Marginalize: Explaining Away Label Noise with Privileged
Information},
booktitle = {International Conference on Machine Learning, {ICML} 2022, 17-23 July
2022, Baltimore, Maryland, {USA} },
series = {Proceedings of Machine Learning Research},
volume = {162},
pages = {4219--4237},
publisher = { {PMLR} },
year = {2022},
url = {https://proceedings.mlr.press/v162/collier22a.html},
}
@article{ILSVRC15,
Author = {Olga Russakovsky and Jia Deng and Hao Su and Jonathan Krause and Sanjeev Satheesh and Sean Ma and Zhiheng Huang and Andrej Karpathy and Aditya Khosla and Michael Bernstein and Alexander C. Berg and Li Fei-Fei},
Title = { {ImageNet Large Scale Visual Recognition Challenge} },
Year = {2015},
journal = {International Journal of Computer Vision (IJCV)},
doi = {10.1007/s11263-015-0816-y},
volume={115},
number={3},
pages={211-252}
}