+
+## YOLO：简史
+
+[YOLO](https://arxiv.org/abs/1506.02640) (You Only Look Once)，由华盛顿大学的Joseph Redmon和Ali Farhadi开发的流行目标检测和图像分割模型，于2015年推出，由于其高速和准确性而迅速流行。
+
+- [YOLOv2](https://arxiv.org/abs/1612.08242) 在2016年发布，通过引入批量归一化、锚框和维度聚类来改进了原始模型。
+- [YOLOv3](https://pjreddie.com/media/files/papers/YOLOv3.pdf) 在2018年推出，进一步增强了模型的性能，使用了更高效的主干网络、多个锚点和空间金字塔池化。
+- [YOLOv4](https://arxiv.org/abs/2004.10934) 在2020年发布，引入了Mosaic数据增强、新的无锚检测头和新的损失函数等创新功能。
+- [YOLOv5](https://github.com/ultralytics/yolov5) 进一步改进了模型的性能，并增加了新功能，如超参数优化、集成实验跟踪和自动导出到常用的导出格式。
+- [YOLOv6](https://github.com/meituan/YOLOv6) 在2022年由[美团](https://about.meituan.com/)开源，现在正在该公司的许多自动送货机器人中使用。
+- [YOLOv7](https://github.com/WongKinYiu/yolov7) 在COCO关键点数据集上添加了额外的任务，如姿态估计。
+- [YOLOv8](https://github.com/ultralytics/ultralytics) 是Ultralytics的YOLO的最新版本。作为一种前沿、最先进(SOTA)的模型，YOLOv8在之前版本的成功基础上引入了新功能和改进，以提高性能、灵活性和效率。YOLOv8支持全范围的视觉AI任务，包括[检测](https://docs.ultralytics.com/tasks/detect/), [分割](https://docs.ultralytics.com/tasks/segment/), [姿态估计](https://docs.ultralytics.com/tasks/pose/), [跟踪](https://docs.ultralytics.com/modes/track/), 和[分类](https://docs.ultralytics.com/tasks/classify/)。这种多功能性使用户能够利用YOLOv8的功能应对多种应用和领域的需求。
+
+## YOLO许可证：Ultralytics YOLO是如何授权的？
+
+Ultralytics提供两种许可选项以适应不同的使用场景：
+
+- **AGPL-3.0许可证**：这种[OSI-approved](https://opensource.org/licenses/)开源许可证非常适合学生和爱好者，促进了开放的合作和知识共享。更多详细信息请参阅[LICENSE](https://github.com/ultralytics/ultralytics/blob/main/LICENSE)文件。
+- **企业许可证**：这种许可证设计用于商业用途，允许将Ultralytics软件和AI模型无缝集成到商业商品和服务中，绕过AGPL-3.0的开源要求。如果您的场景涉及将我们的解决方案嵌入到商业产品中，请通过[Ultralytics Licensing](https://ultralytics.com/license)联系我们。
+
+我们的授权策略旨在确保我们的开源项目的任何改进都能回馈到社区。我们十分珍视开源原则❤️，我们的使命是确保我们的贡献能够以对所有人有益的方式被利用和拓展。
+
+---
+
+**注意**：我们正在努力为我们的文档页面提供中文文档，并希望在接下来的几个月内发布。请密切关注我们的更新，并感谢您的耐心等待。
diff --git a/mkdocs.yml b/mkdocs.yml
index 1bf34dab..848e9733 100644
--- a/mkdocs.yml
+++ b/mkdocs.yml
@@ -11,6 +11,7 @@ remote_name: https://github.com/ultralytics/docs
theme:
name: material
+ language: en
custom_dir: docs/overrides
logo: https://github.com/ultralytics/assets/raw/main/logo/Ultralytics_Logotype_Reverse.svg
favicon: assets/favicon.ico
@@ -25,14 +26,14 @@ theme:
- scheme: default
# primary: grey
toggle:
- icon: material/brightness-7
+ icon: material/toggle-switch
name: Switch to dark mode
# Palette toggle for dark mode
- scheme: slate
# primary: black
toggle:
- icon: material/brightness-4
+ icon: material/toggle-switch-off-outline
name: Switch to light mode
features:
- announce.dismiss
@@ -44,15 +45,17 @@ theme:
- search.share
- search.suggest
- toc.follow
- - toc.integrate
+ # - toc.integrate
- navigation.top
- navigation.tabs
- navigation.tabs.sticky
- - navigation.expand
+ - navigation.prune
- navigation.footer
- navigation.tracking
- navigation.instant
+ - navigation.instant.progress
- navigation.indexes
+ - navigation.sections
- content.tabs.link # all code tabs change simultaneously
# Customization
@@ -64,20 +67,13 @@ extra:
analytics:
provider: google
property: G-2M5EHKC0BH
- # feedback:
- # title: Was this page helpful?
- # ratings:
- # - icon: material/heart
- # name: This page was helpful
- # data: 1
- # note: Thanks for your feedback!
- # - icon: material/heart-broken
- # name: This page could be improved
- # data: 0
- # note: >-
- # Thanks for your feedback!
- # Tell us what we can improve.
-
+ alternate: # language drop-down
+ - name: English
+ link: /
+ lang: en
+ - name: 简体中文
+ link: /zh/
+ lang: zh
social:
- icon: fontawesome/brands/github
link: https://github.com/ultralytics
@@ -145,6 +141,8 @@ nav:
- Segment: tasks/segment.md
- Classify: tasks/classify.md
- Pose: tasks/pose.md
+ - Ultralytics 文档:
+ - zh/index.md
- Quickstart: quickstart.md
- Modes:
- modes/index.md
diff --git a/setup.py b/setup.py
index 3b8d4643..faa4b9c4 100644
--- a/setup.py
+++ b/setup.py
@@ -75,7 +75,7 @@ setup(
'mkdocs-material',
'mkdocstrings[python]',
'mkdocs-redirects', # for 301 redirects
- 'mkdocs-ultralytics-plugin>=0.0.29', # for meta descriptions and images, dates and authors
+ 'mkdocs-ultralytics-plugin>=0.0.30', # for meta descriptions and images, dates and authors
],
'export': [
'coremltools>=7.0',
diff --git a/ultralytics/data/base.py b/ultralytics/data/base.py
index 0762c5fb..1df546b3 100644
--- a/ultralytics/data/base.py
+++ b/ultralytics/data/base.py
@@ -267,20 +267,28 @@ class BaseDataset(Dataset):
return label
def build_transforms(self, hyp=None):
- """Users can custom augmentations here
- like:
+ """
+ Users can customize augmentations here.
+
+ Example:
+ ```python
if self.augment:
# Training transforms
return Compose([])
else:
# Val transforms
return Compose([])
+ ```
"""
raise NotImplementedError
def get_labels(self):
- """Users can custom their own format here.
- Make sure your output is a list with each element like below:
+ """
+ Users can customize their own format here.
+
+ Note:
+ Ensure output is a dictionary with the following keys:
+ ```python
dict(
im_file=im_file,
shape=shape, # format: (height, width)
@@ -291,5 +299,6 @@ class BaseDataset(Dataset):
normalized=True, # or False
bbox_format="xyxy", # or xywh, ltwh
)
+ ```
"""
raise NotImplementedError
diff --git a/ultralytics/data/utils.py b/ultralytics/data/utils.py
index c5c2d994..68876e67 100644
--- a/ultralytics/data/utils.py
+++ b/ultralytics/data/utils.py
@@ -154,28 +154,40 @@ def verify_image_label(args):
def polygon2mask(imgsz, polygons, color=1, downsample_ratio=1):
"""
+ Convert a list of polygons to a binary mask of the specified image size.
+
Args:
- imgsz (tuple): The image size.
- polygons (list[np.ndarray]): [N, M], N is the number of polygons, M is the number of points(Be divided by 2).
- color (int): color
- downsample_ratio (int): downsample ratio
+ imgsz (tuple): The size of the image as (height, width).
+ polygons (list[np.ndarray]): A list of polygons. Each polygon is an array with shape [N, M], where
+ N is the number of polygons, and M is the number of points such that M % 2 = 0.
+ color (int, optional): The color value to fill in the polygons on the mask. Defaults to 1.
+ downsample_ratio (int, optional): Factor by which to downsample the mask. Defaults to 1.
+
+ Returns:
+ (np.ndarray): A binary mask of the specified image size with the polygons filled in.
"""
mask = np.zeros(imgsz, dtype=np.uint8)
polygons = np.asarray(polygons, dtype=np.int32)
polygons = polygons.reshape((polygons.shape[0], -1, 2))
cv2.fillPoly(mask, polygons, color=color)
nh, nw = (imgsz[0] // downsample_ratio, imgsz[1] // downsample_ratio)
- # NOTE: fillPoly first then resize is trying to keep the same way of loss calculation when mask-ratio=1.
+ # Note: fillPoly first then resize is trying to keep the same loss calculation method when mask-ratio=1
return cv2.resize(mask, (nw, nh))
def polygons2masks(imgsz, polygons, color, downsample_ratio=1):
"""
+ Convert a list of polygons to a set of binary masks of the specified image size.
+
Args:
- imgsz (tuple): The image size.
- polygons (list[np.ndarray]): each polygon is [N, M], N is number of polygons, M is number of points (M % 2 = 0)
- color (int): color
- downsample_ratio (int): downsample ratio
+ imgsz (tuple): The size of the image as (height, width).
+ polygons (list[np.ndarray]): A list of polygons. Each polygon is an array with shape [N, M], where
+ N is the number of polygons, and M is the number of points such that M % 2 = 0.
+ color (int): The color value to fill in the polygons on the masks.
+ downsample_ratio (int, optional): Factor by which to downsample each mask. Defaults to 1.
+
+ Returns:
+ (np.ndarray): A set of binary masks of the specified image size with the polygons filled in.
"""
return np.array([polygon2mask(imgsz, [x.reshape(-1)], color, downsample_ratio) for x in polygons])
@@ -205,7 +217,7 @@ def find_dataset_yaml(path: Path) -> Path:
Find and return the YAML file associated with a Detect, Segment or Pose dataset.
This function searches for a YAML file at the root level of the provided directory first, and if not found, it
- performs a recursive search. It prefers YAML files that have the samestem as the provided path. An AssertionError
+ performs a recursive search. It prefers YAML files that have the same stem as the provided path. An AssertionError
is raised if no YAML file is found or if multiple YAML files are found.
Args:
@@ -438,7 +450,8 @@ class HUBDatasetStats:
self.stats = {'nc': len(data['names']), 'names': list(data['names'].values())} # statistics dictionary
self.data = data
- def _unzip(self, path):
+ @staticmethod
+ def _unzip(path):
"""Unzip data.zip."""
if not str(path).endswith('.zip'): # path is data.yaml
return False, None, path
diff --git a/ultralytics/models/sam/modules/decoders.py b/ultralytics/models/sam/modules/decoders.py
index 999b5be3..4ad1d9f1 100644
--- a/ultralytics/models/sam/modules/decoders.py
+++ b/ultralytics/models/sam/modules/decoders.py
@@ -171,7 +171,7 @@ class MLP(nn.Module):
hidden_dim (int): The dimensionality of the hidden layers.
output_dim (int): The dimensionality of the output layer.
num_layers (int): The number of hidden layers.
- sigmoid_output (bool, optional): Whether to apply a sigmoid activation to the output layer. Defaults to False.
+ sigmoid_output (bool, optional): Apply a sigmoid activation to the output layer. Defaults to False.
"""
super().__init__()
self.num_layers = num_layers
diff --git a/ultralytics/models/sam/modules/encoders.py b/ultralytics/models/sam/modules/encoders.py
index 30e6014e..f7771380 100644
--- a/ultralytics/models/sam/modules/encoders.py
+++ b/ultralytics/models/sam/modules/encoders.py
@@ -10,13 +10,15 @@ import torch.nn.functional as F
from ultralytics.nn.modules import LayerNorm2d, MLPBlock
-# This class and its supporting functions below lightly adapted from the ViTDet backbone available at: https://github.com/facebookresearch/detectron2/blob/main/detectron2/modeling/backbone/vit.py # noqa
class ImageEncoderViT(nn.Module):
"""
An image encoder using Vision Transformer (ViT) architecture for encoding an image into a compact latent space. The
encoder takes an image, splits it into patches, and processes these patches through a series of transformer blocks.
The encoded patches are then processed through a neck to generate the final encoded representation.
+ This class and its supporting functions below lightly adapted from the ViTDet backbone available at
+ https://github.com/facebookresearch/detectron2/blob/main/detectron2/modeling/backbone/vit.py.
+
Attributes:
img_size (int): Dimension of input images, assumed to be square.
patch_embed (PatchEmbed): Module for patch embedding.
@@ -410,6 +412,8 @@ class Attention(nn.Module):
input_size: Optional[Tuple[int, int]] = None,
) -> None:
"""
+ Initialize Attention module.
+
Args:
dim (int): Number of input channels.
num_heads (int): Number of attention heads.
@@ -502,8 +506,8 @@ def window_unpartition(windows: torch.Tensor, window_size: int, pad_hw: Tuple[in
def get_rel_pos(q_size: int, k_size: int, rel_pos: torch.Tensor) -> torch.Tensor:
"""
- Get relative positional embeddings according to the relative positions of
- query and key sizes.
+ Get relative positional embeddings according to the relative positions of query and key sizes.
+
Args:
q_size (int): size of query q.
k_size (int): size of key k.
@@ -542,8 +546,9 @@ def add_decomposed_rel_pos(
k_size: Tuple[int, int],
) -> torch.Tensor:
"""
- Calculate decomposed Relative Positional Embeddings from :paper:`mvitv2`.
- https://github.com/facebookresearch/mvit/blob/19786631e330df9f3622e5402b4a419a263a2c80/mvit/models/attention.py # noqa B950
+ Calculate decomposed Relative Positional Embeddings from mvitv2 paper at
+ https://github.com/facebookresearch/mvit/blob/main/mvit/models/attention.py.
+
Args:
attn (Tensor): attention map.
q (Tensor): query q in the attention layer with shape (B, q_h * q_w, C).
@@ -583,6 +588,8 @@ class PatchEmbed(nn.Module):
embed_dim: int = 768,
) -> None:
"""
+ Initialize PatchEmbed module.
+
Args:
kernel_size (Tuple): kernel size of the projection layer.
stride (Tuple): stride of the projection layer.
diff --git a/ultralytics/nn/modules/transformer.py b/ultralytics/nn/modules/transformer.py
index 493ca3c6..9fe9597f 100644
--- a/ultralytics/nn/modules/transformer.py
+++ b/ultralytics/nn/modules/transformer.py
@@ -39,7 +39,8 @@ class TransformerEncoderLayer(nn.Module):
self.act = act
self.normalize_before = normalize_before
- def with_pos_embed(self, tensor, pos=None):
+ @staticmethod
+ def with_pos_embed(tensor, pos=None):
"""Add position embeddings to the tensor if provided."""
return tensor if pos is None else tensor + pos
@@ -180,9 +181,10 @@ class LayerNorm2d(nn.Module):
"""
2D Layer Normalization module inspired by Detectron2 and ConvNeXt implementations.
- Original implementation at
+ Original implementations in
https://github.com/facebookresearch/detectron2/blob/main/detectron2/layers/batch_norm.py
- https://github.com/facebookresearch/ConvNeXt/blob/d1fa8f6fef0a165b27399986cc2bdacc92777e40/models/convnext.py#L119
+ and
+ https://github.com/facebookresearch/ConvNeXt/blob/main/models/convnext.py.
"""
def __init__(self, num_channels, eps=1e-6):
@@ -250,7 +252,7 @@ class MSDeformAttn(nn.Module):
def forward(self, query, refer_bbox, value, value_shapes, value_mask=None):
"""
- Perform forward pass for multi-scale deformable attention.
+ Perform forward pass for multiscale deformable attention.
https://github.com/PaddlePaddle/PaddleDetection/blob/develop/ppdet/modeling/transformers/deformable_transformer.py
diff --git a/ultralytics/utils/metrics.py b/ultralytics/utils/metrics.py
index 16f8019f..761a71d3 100644
--- a/ultralytics/utils/metrics.py
+++ b/ultralytics/utils/metrics.py
@@ -48,8 +48,7 @@ def bbox_ioa(box1, box2, iou=False, eps=1e-7):
def box_iou(box1, box2, eps=1e-7):
"""
- Calculate intersection-over-union (IoU) of boxes.
- Both sets of boxes are expected to be in (x1, y1, x2, y2) format.
+ Calculate intersection-over-union (IoU) of boxes. Both sets of boxes are expected to be in (x1, y1, x2, y2) format.
Based on https://github.com/pytorch/vision/blob/master/torchvision/ops/boxes.py
Args:
diff --git a/ultralytics/utils/tal.py b/ultralytics/utils/tal.py
index 1d10b7aa..8a8333f1 100644
--- a/ultralytics/utils/tal.py
+++ b/ultralytics/utils/tal.py
@@ -61,8 +61,8 @@ class TaskAlignedAssigner(nn.Module):
"""
A task-aligned assigner for object detection.
- This class assigns ground-truth (gt) objects to anchors based on the task-aligned metric,
- which combines both classification and localization information.
+ This class assigns ground-truth (gt) objects to anchors based on the task-aligned metric, which combines both
+ classification and localization information.
Attributes:
topk (int): The number of top candidates to consider.
@@ -85,8 +85,8 @@ class TaskAlignedAssigner(nn.Module):
@torch.no_grad()
def forward(self, pd_scores, pd_bboxes, anc_points, gt_labels, gt_bboxes, mask_gt):
"""
- Compute the task-aligned assignment.
- Reference https://github.com/Nioolek/PPYOLOE_pytorch/blob/master/ppyoloe/assigner/tal_assigner.py
+ Compute the task-aligned assignment. Reference code is available at
+ https://github.com/Nioolek/PPYOLOE_pytorch/blob/master/ppyoloe/assigner/tal_assigner.py.
Args:
pd_scores (Tensor): shape(bs, num_total_anchors, num_classes)
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+