HongChuang/ultralytics-ascend
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 3

ultralytics 8.0.239 Ultralytics Actions and hub-sdk adoption (#7431)

Browse source 
Signed-off-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: UltralyticsAssistant <web@ultralytics.com>
Co-authored-by: Burhan <62214284+Burhan-Q@users.noreply.github.com>
Co-authored-by: Kayzwer <68285002+Kayzwer@users.noreply.github.com>
This commit is contained in:
Glenn Jocher 2024-01-10 03:16:08 +01:00 • committed by GitHub
parent e795277391
commit fe27db2f6e
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
139 changed files with 6870 additions and 5125 deletions
Download patch file Download diff file Expand all files Collapse all files

108
ultralytics/nn/modules/head.py
View file

@ -14,11 +14,12 @@ from .conv import Conv
from .transformer import MLP, DeformableTransformerDecoder, DeformableTransformerDecoderLayer
from .utils import bias_init_with_prob, linear_init_
__all__ = 'Detect', 'Segment', 'Pose', 'Classify', 'OBB', 'RTDETRDecoder'
__all__ = "Detect", "Segment", "Pose", "Classify", "OBB", "RTDETRDecoder"
class Detect(nn.Module):
"""YOLOv8 Detect head for detection models."""
dynamic = False # force grid reconstruction
export = False # export mode
shape = None
@ -35,7 +36,8 @@ class Detect(nn.Module):
self.stride = torch.zeros(self.nl) # strides computed during build
c2, c3 = max((16, ch[0] // 4, self.reg_max * 4)), max(ch[0], min(self.nc, 100)) # channels
self.cv2 = nn.ModuleList(
nn.Sequential(Conv(x, c2, 3), Conv(c2, c2, 3), nn.Conv2d(c2, 4 * self.reg_max, 1)) for x in ch)
nn.Sequential(Conv(x, c2, 3), Conv(c2, c2, 3), nn.Conv2d(c2, 4 * self.reg_max, 1)) for x in ch
)
self.cv3 = nn.ModuleList(nn.Sequential(Conv(x, c3, 3), Conv(c3, c3, 3), nn.Conv2d(c3, self.nc, 1)) for x in ch)
self.dfl = DFL(self.reg_max) if self.reg_max > 1 else nn.Identity()
@ -53,14 +55,14 @@ class Detect(nn.Module):
self.anchors, self.strides = (x.transpose(0, 1) for x in make_anchors(x, self.stride, 0.5))
self.shape = shape
if self.export and self.format in ('saved_model', 'pb', 'tflite', 'edgetpu', 'tfjs'): # avoid TF FlexSplitV ops
box = x_cat[:, :self.reg_max * 4]
cls = x_cat[:, self.reg_max * 4:]
if self.export and self.format in ("saved_model", "pb", "tflite", "edgetpu", "tfjs"): # avoid TF FlexSplitV ops
box = x_cat[:, : self.reg_max * 4]
cls = x_cat[:, self.reg_max * 4 :]
else:
box, cls = x_cat.split((self.reg_max * 4, self.nc), 1)
dbox = self.decode_bboxes(box)
if self.export and self.format in ('tflite', 'edgetpu'):
if self.export and self.format in ("tflite", "edgetpu"):
# Precompute normalization factor to increase numerical stability
# See https://github.com/ultralytics/ultralytics/issues/7371
img_h = shape[2]
@ -79,7 +81,7 @@ class Detect(nn.Module):
# ncf = math.log(0.6 / (m.nc - 0.999999)) if cf is None else torch.log(cf / cf.sum()) # nominal class frequency
for a, b, s in zip(m.cv2, m.cv3, m.stride): # from
a[-1].bias.data[:] = 1.0 # box
b[-1].bias.data[:m.nc] = math.log(5 / m.nc / (640 / s) ** 2) # cls (.01 objects, 80 classes, 640 img)
b[-1].bias.data[: m.nc] = math.log(5 / m.nc / (640 / s) ** 2) # cls (.01 objects, 80 classes, 640 img)
def decode_bboxes(self, bboxes):
"""Decode bounding boxes."""
@ -214,26 +216,28 @@ class RTDETRDecoder(nn.Module):
and class labels for objects in an image. It integrates features from multiple layers and runs through a series of
Transformer decoder layers to output the final predictions.
"""
export = False # export mode
def __init__(
self,
nc=80,
ch=(512, 1024, 2048),
hd=256, # hidden dim
nq=300, # num queries
ndp=4, # num decoder points
nh=8, # num head
ndl=6, # num decoder layers
d_ffn=1024, # dim of feedforward
dropout=0.,
act=nn.ReLU(),
eval_idx=-1,
# Training args
nd=100, # num denoising
label_noise_ratio=0.5,
box_noise_scale=1.0,
learnt_init_query=False):
self,
nc=80,
ch=(512, 1024, 2048),
hd=256, # hidden dim
nq=300, # num queries
ndp=4, # num decoder points
nh=8, # num head
ndl=6, # num decoder layers
d_ffn=1024, # dim of feedforward
dropout=0.0,
act=nn.ReLU(),
eval_idx=-1,
# Training args
nd=100, # num denoising
label_noise_ratio=0.5,
box_noise_scale=1.0,
learnt_init_query=False,
):
"""
Initializes the RTDETRDecoder module with the given parameters.
@ -302,28 +306,30 @@ class RTDETRDecoder(nn.Module):
feats, shapes = self._get_encoder_input(x)
# Prepare denoising training
dn_embed, dn_bbox, attn_mask, dn_meta = \
get_cdn_group(batch,
self.nc,
self.num_queries,
self.denoising_class_embed.weight,
self.num_denoising,
self.label_noise_ratio,
self.box_noise_scale,
self.training)
dn_embed, dn_bbox, attn_mask, dn_meta = get_cdn_group(
batch,
self.nc,
self.num_queries,
self.denoising_class_embed.weight,
self.num_denoising,
self.label_noise_ratio,
self.box_noise_scale,
self.training,
)
embed, refer_bbox, enc_bboxes, enc_scores = \
self._get_decoder_input(feats, shapes, dn_embed, dn_bbox)
embed, refer_bbox, enc_bboxes, enc_scores = self._get_decoder_input(feats, shapes, dn_embed, dn_bbox)
# Decoder
dec_bboxes, dec_scores = self.decoder(embed,
refer_bbox,
feats,
shapes,
self.dec_bbox_head,
self.dec_score_head,
self.query_pos_head,
attn_mask=attn_mask)
dec_bboxes, dec_scores = self.decoder(
embed,
refer_bbox,
feats,
shapes,
self.dec_bbox_head,
self.dec_score_head,
self.query_pos_head,
attn_mask=attn_mask,
)
x = dec_bboxes, dec_scores, enc_bboxes, enc_scores, dn_meta
if self.training:
return x
@ -331,24 +337,24 @@ class RTDETRDecoder(nn.Module):
y = torch.cat((dec_bboxes.squeeze(0), dec_scores.squeeze(0).sigmoid()), -1)
return y if self.export else (y, x)
def _generate_anchors(self, shapes, grid_size=0.05, dtype=torch.float32, device='cpu', eps=1e-2):
def _generate_anchors(self, shapes, grid_size=0.05, dtype=torch.float32, device="cpu", eps=1e-2):
"""Generates anchor bounding boxes for given shapes with specific grid size and validates them."""
anchors = []
for i, (h, w) in enumerate(shapes):
sy = torch.arange(end=h, dtype=dtype, device=device)
sx = torch.arange(end=w, dtype=dtype, device=device)
grid_y, grid_x = torch.meshgrid(sy, sx, indexing='ij') if TORCH_1_10 else torch.meshgrid(sy, sx)
grid_y, grid_x = torch.meshgrid(sy, sx, indexing="ij") if TORCH_1_10 else torch.meshgrid(sy, sx)
grid_xy = torch.stack([grid_x, grid_y], -1) # (h, w, 2)
valid_WH = torch.tensor([w, h], dtype=dtype, device=device)
grid_xy = (grid_xy.unsqueeze(0) + 0.5) / valid_WH # (1, h, w, 2)
wh = torch.ones_like(grid_xy, dtype=dtype, device=device) * grid_size * (2.0 ** i)
wh = torch.ones_like(grid_xy, dtype=dtype, device=device) * grid_size * (2.0**i)
anchors.append(torch.cat([grid_xy, wh], -1).view(-1, h * w, 4)) # (1, h*w, 4)
anchors = torch.cat(anchors, 1) # (1, h*w*nl, 4)
valid_mask = ((anchors > eps) * (anchors < 1 - eps)).all(-1, keepdim=True) # 1, h*w*nl, 1
anchors = torch.log(anchors / (1 - anchors))
anchors = anchors.masked_fill(~valid_mask, float('inf'))
anchors = anchors.masked_fill(~valid_mask, float("inf"))
return anchors, valid_mask
def _get_encoder_input(self, x):
@ -415,13 +421,13 @@ class RTDETRDecoder(nn.Module):
# NOTE: the weight initialization in `linear_init_` would cause NaN when training with custom datasets.
# linear_init_(self.enc_score_head)
constant_(self.enc_score_head.bias, bias_cls)
constant_(self.enc_bbox_head.layers[-1].weight, 0.)
constant_(self.enc_bbox_head.layers[-1].bias, 0.)
constant_(self.enc_bbox_head.layers[-1].weight, 0.0)
constant_(self.enc_bbox_head.layers[-1].bias, 0.0)
for cls_, reg_ in zip(self.dec_score_head, self.dec_bbox_head):
# linear_init_(cls_)
constant_(cls_.bias, bias_cls)
constant_(reg_.layers[-1].weight, 0.)
constant_(reg_.layers[-1].bias, 0.)
constant_(reg_.layers[-1].weight, 0.0)
constant_(reg_.layers[-1].bias, 0.0)
linear_init_(self.enc_output[0])
xavier_uniform_(self.enc_output[0].weight)