- คำอธิบาย :
ดาวน์โหลด SA-1B
Segment Anything 1 Billion (SA-1B) เป็นชุดข้อมูลที่ออกแบบมาเพื่อฝึกโมเดลการแบ่งส่วนวัตถุสำหรับวัตถุประสงค์ทั่วไปจากภาพโลกเปิด ชุดข้อมูลได้รับการแนะนำในรายงาน "Segment Anything"
ชุดข้อมูล SA-1B ประกอบด้วยรูปภาพที่หลากหลาย มีความละเอียดสูง มีลิขสิทธิ์ และปกป้องความเป็นส่วนตัวจำนวน 11 ล้านภาพ และคำอธิบายประกอบแบบมาสก์ 1.1B มาสก์มีให้ในรูปแบบการเข้ารหัสความยาวรัน COCO (RLE) และไม่มีคลาส
ใบอนุญาตเป็นแบบกำหนดเอง โปรดอ่านข้อกำหนดและเงื่อนไขฉบับเต็มได้ที่ https://ai.facebook.com/datasets/segment-anything-downloads
คุณสมบัติทั้งหมดอยู่ในชุดข้อมูลดั้งเดิม ยกเว้น image.content (เนื้อหาของรูปภาพ)
คุณสามารถถอดรหัสมาสก์การแบ่งเซ็กเมนต์ได้ด้วย:
import tensorflow_datasets as tfds
pycocotools = tfds.core.lazy_imports.pycocotools
ds = tfds.load('segment_anything', split='train')
for example in tfds.as_numpy(ds):
segmentation = example['annotations']['segmentation']
for counts, size in zip(segmentation['counts'], segmentation['size']):
encoded_mask = {'size': size, 'counts': counts}
mask = pycocotools.decode(encoded_mask) # np.array(dtype=uint8) mask
...
หน้าแรก : https://ai.facebook.com/datasets/segment-anything-downloads
ซอร์สโค้ด :
tfds.datasets.segment_anything.Builderรุ่น :
-
1.0.0(ค่าเริ่มต้น): การเปิดตัวครั้งแรก
-
ขนาดดาวน์โหลด :
10.28 TiBขนาดชุดข้อมูล :
10.59 TiBคำแนะนำในการดาวน์โหลดด้วยตนเอง : ชุดข้อมูลนี้ต้องการให้คุณดาวน์โหลดข้อมูลต้นฉบับด้วยตนเองลงใน
download_config.manual_dir(ค่าเริ่มต้นเป็น~/tensorflow_datasets/downloads/manual/):
ดาวน์โหลดไฟล์ลิงก์จาก https://ai.facebook.com/datasets/segment-anything-downloadsmanual_dirควรมีไฟล์ลิงก์ที่บันทึกเป็นเซ็กเมนต์_anything_links.txtแคชอัตโนมัติ ( เอกสาร ): No
แยก :
| แยก | ตัวอย่าง |
|---|---|
'train' | 11,185,362 |
- โครงสร้างคุณสมบัติ :
FeaturesDict({
'annotations': Sequence({
'area': Scalar(shape=(), dtype=uint64, description=The area in pixels of the mask.),
'bbox': BBoxFeature(shape=(4,), dtype=float32, description=The box around the mask, in TFDS format.),
'crop_box': BBoxFeature(shape=(4,), dtype=float32, description=The crop of the image used to generate the mask, in TFDS format.),
'id': Scalar(shape=(), dtype=uint64, description=Identifier for the annotation.),
'point_coords': Tensor(shape=(1, 2), dtype=float64, description=The point coordinates input to the model to generate the mask.),
'predicted_iou': Scalar(shape=(), dtype=float64, description=The model's own prediction of the mask's quality.),
'segmentation': FeaturesDict({
'counts': string,
'size': Tensor(shape=(2,), dtype=uint64),
}),
'stability_score': Scalar(shape=(), dtype=float64, description=A measure of the mask's quality.),
}),
'image': FeaturesDict({
'content': Image(shape=(None, None, 3), dtype=uint8, description=Content of the image.),
'file_name': string,
'height': uint64,
'image_id': uint64,
'width': uint64,
}),
})
- เอกสารคุณสมบัติ :
| คุณสมบัติ | ระดับ | รูปร่าง | ประเภทD | คำอธิบาย |
|---|---|---|---|---|
| คุณสมบัติDict | ||||
| คำอธิบายประกอบ | ลำดับ | |||
| คำอธิบายประกอบ/พื้นที่ | สเกลาร์ | uint64 | พื้นที่เป็นพิกเซลของมาสก์ | |
| คำอธิบายประกอบ/bbox | คุณสมบัติ BBox | (4,) | ลอย32 | กล่องรอบมาสก์ ในรูปแบบ TFDS |
| คำอธิบายประกอบ/crop_box | คุณสมบัติ BBox | (4,) | ลอย32 | การครอบตัดรูปภาพที่ใช้สร้างมาสก์ ในรูปแบบ TFDS |
| คำอธิบายประกอบ/รหัส | สเกลาร์ | uint64 | ตัวระบุสำหรับคำอธิบายประกอบ | |
| คำอธิบายประกอบ/point_coords | เทนเซอร์ | (1, 2) | ลอย64 | พิกัดจุดอินพุตไปยังโมเดลเพื่อสร้างมาสก์ |
| คำอธิบายประกอบ/predicted_iou | สเกลาร์ | ลอย64 | การคาดการณ์คุณภาพของหน้ากากของตัวแบบเอง | |
| คำอธิบายประกอบ/การแบ่งส่วน | คุณสมบัติDict | มาสก์การแบ่งส่วนที่เข้ารหัสในรูปแบบ COCO RLE (กำหนดด้วย size คีย์และ counts ) | ||
| คำอธิบายประกอบ/การแบ่งส่วน/การนับ | เทนเซอร์ | เชือก | ||
| คำอธิบายประกอบ/การแบ่งส่วน/ขนาด | เทนเซอร์ | (2,) | uint64 | |
| คำอธิบายประกอบ/stability_score | สเกลาร์ | ลอย64 | การวัดคุณภาพของหน้ากาก | |
| ภาพ | คุณสมบัติDict | |||
| รูปภาพ/เนื้อหา | ภาพ | (ไม่มี ไม่มี 3) | uint8 | เนื้อหาของภาพ |
| รูปภาพ/file_name | เทนเซอร์ | เชือก | ||
| รูปภาพ/ส่วนสูง | เทนเซอร์ | uint64 | ||
| รูปภาพ/image_id | เทนเซอร์ | uint64 | ||
| รูปภาพ/ความกว้าง | เทนเซอร์ | uint64 |
คีย์ภายใต้การดูแล (ดู
as_superviseddoc ):Noneรูปภาพ ( tfds.show_examples ): ไม่รองรับ
ตัวอย่าง ( tfds.as_dataframe ):
- การอ้างอิง :
@misc{kirillov2023segment,
title={Segment Anything},
author={Alexander Kirillov and Eric Mintun and Nikhila Ravi and Hanzi Mao and Chloe Rolland and Laura Gustafson and Tete Xiao and Spencer Whitehead and Alexander C. Berg and Wan-Yen Lo and Piotr Dollár and Ross Girshick},
year={2023},
eprint={2304.02643},
archivePrefix={arXiv},
primaryClass={cs.CV}
}