Skip to content

Add YOLO11-Pose P2 config #63

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Y-T-G merged 1 commit into from
Apr 26, 2025
Merged

Add YOLO11-Pose P2 config #63

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
87 changes: 87 additions & 0 deletions cfg/pose/yolo11-p2.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,87 @@
# Metadata
author: Y-T-G
task: pose
keywords: [cnn, yolo11, small-object]
description: YOLO11 keypoint estimation model with an extra P2 scale for small object detecton.
flops:
n: 14.6
s: 34.3
m: 96.2
l: 120.8
x: 255.1
parameters:
n: 3065412
s: 10171668
m: 21342452
l: 26867700
x: 58923924
min_version: 8.3.0

# Model
nc: 80
strides: [4, 8, 16, 32]
kpt_shape: [17, 3]
scales:
n: [0.50, 0.25, 1024]
s: [0.50, 0.50, 1024]
m: [0.50, 1.00, 512]
l: [1.00, 1.00, 512]
x: [1.00, 1.50, 512]

backbone:
# Stem
- [-1, 1, Conv, [64, 3, 2]] # (1, 64, 320, 320) - 0

# P2 Backbone
- [-1, 1, Conv, [128, 3, 2]] # (1, 128, 160, 160) - 1
- [-1, 2, C3k2, [256, False, 0.25]] # (1, 256, 160, 160) - 2

# P3 Backbone
- [-1, 1, Conv, [256, 3, 2]] # (1, 256, 80, 80) - 3
- [-1, 2, C3k2, [512, False, 0.25]] # (1, 512, 80, 80) - 4

# P4 Backbone
- [-1, 1, Conv, [512, 3, 2]] # (1, 512, 40, 40) - 5
- [-1, 2, C3k2, [512, True]] # (1, 512, 40, 40) - 6

# P5 Backbone
- [-1, 1, Conv, [1024, 3, 2]] # (1, 512, 20, 20) - 7
- [-1, 2, C3k2, [1024, True]] # (1, 512, 20, 20) - 8
- [-1, 1, SPPF, [1024, 5]] # (1, 512, 20, 20) - 9
- [-1, 2, C2PSA, [1024]] # (1, 512, 20, 20) - 10

head:
# Top-Down Path (Feature Pyramid)
# P4 Head
- [-1, 1, nn.Upsample, [None, 2, nearest]] # (1, 512, 40, 40) - 11
- [[-1, 6], 1, Concat, [1]] # (1, 1024, 40, 40) - 12
- [-1, 2, C3k2, [512, False]] # (1, 512, 40, 40) - 13

# P3 Head
- [-1, 1, nn.Upsample, [None, 2, nearest]] # (1, 512, 80, 80) - 14
- [[-1, 4], 1, Concat, [1]] # (1, 1024, 80, 80) - 15
- [-1, 2, C3k2, [256, False]] # (1, 256, 80, 80) - 16

# P2 Head
- [-1, 1, nn.Upsample, [None, 2, nearest]] # (1, 256, 160, 160) - 17
- [[-1, 2], 1, Concat, [1]] # (1, 512, 160, 160) - 18
- [-1, 2, C3k2, [128, False]] # (1, 128, 160, 160) - 19

# Bottom-Up Path (PAN-FPN)
# P3
- [-1, 1, Conv, [128, 3, 2]] # (1, 128, 80, 80) - 20
- [[-1, 16], 1, Concat, [1]] # (1, 384, 80, 80) - 21
- [-1, 2, C3k2, [256, False]] # (1, 256, 80, 80) - 22

# P4
- [-1, 1, Conv, [256, 3, 2]] # (1, 256, 40, 40) - 23
- [[-1, 13], 1, Concat, [1]] # (1, 768, 40, 40) - 24
- [-1, 2, C3k2, [512, False]] # (1, 512, 40, 40) - 25

# P5
- [-1, 1, Conv, [512, 3, 2]] # (1, 512, 20, 20) - 26
- [[-1, 10], 1, Concat, [1]] # (1, 1024, 20, 20) - 27
- [-1, 2, C3k2, [1024, True]] # (1, 512, 20, 20) - 28

# Pose Head (P2, P3, P4, P5)
- [[19, 22, 25, 28], 1, Pose, [nc, kpt_shape]] # 29