Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
H
HardwareAcceleratedPoseTracking
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Wiki
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
Build
Pipelines
Jobs
Pipeline schedules
Artifacts
Deploy
Releases
Package registry
Container registry
Model registry
Operate
Environments
Terraform modules
Monitor
Incidents
Analyze
Value stream analytics
Contributor analytics
CI/CD analytics
Repository analytics
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Terms and privacy
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
Marius Baumann
HardwareAcceleratedPoseTracking
Commits
1ca35140
Commit
1ca35140
authored
3 years ago
by
Michael Schmid
Browse files
Options
Downloads
Plain Diff
Merge branch 'main' of
https://github.com/Nunigan/HardwareAcceleratedPoseTracking
into main
parents
7d01c35b
ddff742b
No related branches found
No related tags found
No related merge requests found
Changes
2
Hide whitespace changes
Inline
Side-by-side
Showing
2 changed files
README.md
+9
-9
9 additions, 9 deletions
README.md
trained_model/ICAIPose/models/Prelu/ICAIPose_5/config.toml
+0
-36
0 additions, 36 deletions
trained_model/ICAIPose/models/Prelu/ICAIPose_5/config.toml
with
9 additions
and
45 deletions
README.md
+
9
−
9
View file @
1ca35140
...
...
@@ -5,17 +5,17 @@ with Xilinx.
## Project: [Hardware Accelerated Pose Tracking](https://www.hackster.io/michi_michi/hardware-accelerated-pose-tracking-d5ebb9)
-
Vitis AI folder
-
Prebuilt
-
All the files needed to run ICAIPose and the hourglass example on the KV260
-
prebuilt.tar.xz the same folder compressed for an easy download directly on the KV260
-
Vitis AI
-
All data to compile the neural network with Vitis AI
-
r
un the start docker.sh file to get the correct
v
itisAI version
.
-
w
ith the script "run_all.sh" all the scripts to compile the network are executed
.
-
h
ourglass
folder
-
The hourglass network from VitisAI Model Zoo
-
w
eights
folder
-
R
un the start docker.sh file to get the correct
V
itis
AI version
-
W
ith the script "run_all.sh" all the scripts
rquired
to compile the network are executed
-
H
ourglass
-
The hourglass network from Vitis
AI Model Zoo
-
W
eights
-
The pretrained weights for the network ICAIPose
-
prebuilt folder
-
all the files needed to run ICAIPose and the hourglass example on the KV260
-
prebuilt.tar.xz the same folder compressed for an easy download to the KV260
-
run_ICAIPose.py
-
Python script with various utility functions
This diff is collapsed.
Click to expand it.
trained_model/ICAIPose/models/Prelu/ICAIPose_5/config.toml
deleted
100644 → 0
+
0
−
36
View file @
7d01c35b
title
=
"Config file ICAIPose"
[project]
name
=
"PoseEstimationTrain"
description
=
"Efficient human pose estimation using modified OpenPose network"
version
=
"1.0.0"
author
=
"Simon Walser"
[network]
width_1
=
[
96
,
256
]
#[96, 256]
width_2
=
[
96
,
256
]
#[128,512]
width_3
=
[
96
,
256
]
#[96, 256]
width_4
=
[
96
,
256
]
#[128,512]
paf_channels
=
28
conf_channels
=
16
[training]
img_res
=
256
epochs
=
800
lr_boundaries
=
[
1
,
200
,
400
,
600
]
# In epochs -> it is converted to steps afterwards
lr_values
=
[
1e-5
,
1e-4
,
5e-5
,
1e-5
,
5e-6
]
# One more than lr_boundaries
evaluate_every
=
541
# In steps
epoch_list
=
[
200
,
400
,
600
]
sigma_list
=
[
13.0
,
10.0
,
7.0
,
5.0
]
loss_factors
=
[
85.0
,
85.0
,
30.2
,
30.2
,
29.7
]
[data]
src_path
=
'/mnt/data/s1walser/PhysioAI/Data/poseestimationtrain/256'
dst_path
=
'/mnt/data/s1walser/PhysioAI/trained_models/poseestimationtrain/256'
# src_path = '/home/simon/work/work_OST/PhysioAI/Data/poseestimationtrain/256'
# dst_path = '/home/simon/work/work_OST/PhysioAI/trained_models/poseestimationtrain/256'
save_attribute
=
'5'
batch_size
=
32
ds_size_train
=
51958
ds_size_valid
=
2142
\ No newline at end of file
This diff is collapsed.
Click to expand it.
Preview
0%
Loading
Try again
or
attach a new file
.
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Save comment
Cancel
Please
register
or
sign in
to comment