Skip to content

Commit c5e040c

Browse files
RoadonePRoadoneP
committed
[21/11/23] first commit
0 parents  commit c5e040c

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

56 files changed

+12656
-0
lines changed

‎.gitignore

+161
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,161 @@
1+
# Byte-compiled / optimized / DLL files
2+
__pycache__/
3+
*.py[cod]
4+
*$py.class
5+
6+
# C extensions
7+
*.so
8+
9+
# Distribution / packaging
10+
.Python
11+
env/
12+
build/
13+
develop-eggs/
14+
dist/
15+
downloads/
16+
eggs/
17+
.eggs/
18+
lib/
19+
lib64/
20+
parts/
21+
sdist/
22+
var/
23+
*.egg-info/
24+
.installed.cfg
25+
*.egg
26+
27+
# PyInstaller
28+
# Usually these files are written by a python script from a template
29+
# before PyInstaller builds the exe, so as to inject date/other infos into it.
30+
*.manifest
31+
*.spec
32+
33+
# Installer logs
34+
pip-log.txt
35+
pip-delete-this-directory.txt
36+
37+
# Unit test / coverage reports
38+
htmlcov/
39+
.tox/
40+
.coverage
41+
.coverage.*
42+
.cache
43+
nosetests.xml
44+
coverage.xml
45+
*,cover
46+
.hypothesis/
47+
48+
# Translations
49+
*.mo
50+
*.pot
51+
52+
# Django stuff:
53+
*.log
54+
local_settings.py
55+
56+
# Flask stuff:
57+
instance/
58+
.webassets-cache
59+
60+
# Scrapy stuff:
61+
.scrapy
62+
63+
# Sphinx documentation
64+
docs/_build/
65+
66+
# PyBuilder
67+
target/
68+
69+
# IPython Notebook
70+
.ipynb_checkpoints
71+
72+
# pyenv
73+
.python-version
74+
75+
# celery beat schedule file
76+
celerybeat-schedule
77+
78+
# dotenv
79+
.env
80+
81+
# virtualenv
82+
venv/
83+
ENV/
84+
85+
# Spyder project settings
86+
.spyderproject
87+
88+
# Rope project settings
89+
.ropeproject
90+
91+
# atom remote-sync package
92+
.remote-sync.json
93+
94+
# weights
95+
weights/
96+
97+
#DS_Store
98+
.DS_Store
99+
100+
# dev stuff
101+
eval/
102+
eval.ipynb
103+
dev.ipynb
104+
.vscode/
105+
106+
# not ready
107+
videos/
108+
templates/
109+
data/ssd_dataloader.py
110+
data/datasets/
111+
doc/visualize.py
112+
read_results.py
113+
ssd300_120000/
114+
demos/live
115+
webdemo.py
116+
test_data_aug.py
117+
118+
# attributes
119+
120+
# pycharm
121+
.idea/
122+
123+
# temp checkout soln
124+
data/coco
125+
data/sbd
126+
data/cityscapes
127+
128+
# pylint
129+
.pylintrc
130+
131+
# ssd.pytorch master branch (for merging)
132+
ssd.pytorch/
133+
134+
# some datasets
135+
data/VOCdevkit/
136+
data/coco/images/
137+
data/coco/annotations/
138+
ap_data.pkl
139+
results/
140+
logs/
141+
scripts/aws/
142+
scripts/gt.npy
143+
scripts/proto.npy
144+
scripts/info.txt
145+
test.pkl
146+
testeval.py
147+
scripts/aws2/
148+
status.sh
149+
train.sh
150+
img/
151+
scripts/aws-ohio/
152+
scripts/aws3/
153+
data/config_dev.py
154+
data/coco/
155+
data/sbd/
156+
157+
vid/
158+
vidres/
159+
160+
_.py
161+
static/

‎1.jpg

64 KB
Loading

‎CHANGELOG.md

+57
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,57 @@
1+
# YOLACT Change Log
2+
3+
This document will detail all changes I make.
4+
I don't know how I'm going to be versioning things yet, so you get dates for now.
5+
6+
```
7+
2020.01.25:
8+
- Fixed the mask IoU branch crashing when all masks in a batch are discarded (fixes #302, #259).
9+
2020.01.24:
10+
- Fixed the conv layer detection during initialization to work with pytorch 1.4 (fixes #292).
11+
2020.01.23:
12+
- Fixed the video playback crashing if there's nothing in the scene (fixes #266).
13+
- Fixed the logger logging the last loss as total loss instead of the actual total (fixes #254).
14+
15+
2019.12.16 (v1.2):
16+
- Added YOLACT++ implementation, paper, and code.
17+
- Added DCN support (need to compile CUDA kernels if you want to use them, see README).
18+
- Added a mask rescoring network trained with mask iou.
19+
- Added configs with more anchors.
20+
21+
2019.12.06:
22+
- Made training much more stable (no more infs and hopefully fewer loss explosions) by ignoring
23+
augmented boxes with < 4px of height and width (this includes 0 area boxes which caused the inf).
24+
See #222 for details.
25+
26+
2019.11.20:
27+
- Fixed bug where saving videos wouldn't work when using cv2 not compiled with display support (#197).
28+
29+
2019.11.06:
30+
- Changed Cython import to only active when using traditional nms.
31+
- Added cross-class fast NMS.
32+
33+
2019.11.04:
34+
- Fixed a bug where the learning rate auto-scaling wasn't being applied properly.
35+
- Fixed a logging bug were lr was sometimes not properly logged after a resume.
36+
37+
2019.10.25 (v1.1):
38+
- Added proper Multi-GPU support. Simply increase your batch size to 8*num_gpus and everything will scale.
39+
- I get an ~1.8x training speed increase when using 2 gpus and an ~3x increase when using 4.
40+
- Added a logger that logs everything about your training.
41+
- Check the Logging section of the README to see how to visualize your logs. (Not written yet)
42+
- Savevideo now uses the evalvideo framework and suports --video_multiframe. It's much faster now!
43+
- Added the ability to display fps right on the videos themselves by using --display_fps
44+
- Evalvideo now doesn't crash when it runs out of frames.
45+
- Pascal SBD is now officially supported! Check the training section for more details.
46+
- Preserve_aspect_ratio kinda sorta works now, but it's iffy and the way I have it set up doesn't perform better.
47+
- Added a ton of new config settings, most of which don't improve performance :/
48+
49+
2019.09.20
50+
- Fixed a bug where custom label maps weren't being applied properly because of global default argument initialization.
51+
2019.08.29
52+
- Fixed a bug where the fpn conv layers weren't getting initialized with xavier since they were being overwritten by jit modules (see #127).
53+
2019.08.04
54+
- Improved the matching algorithm used to match anchors to gt by making it less greedy (see #104).
55+
2019.06.27
56+
- Sped up save video by ~8 ms per frame because I forgot to apply a speed fix I applied to the other modes.
57+
```

‎LICENSE

+21
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,21 @@
1+
MIT License
2+
3+
Copyright (c) 2019 Daniel Bolya
4+
5+
Permission is hereby granted, free of charge, to any person obtaining a copy
6+
of this software and associated documentation files (the "Software"), to deal
7+
in the Software without restriction, including without limitation the rights
8+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
9+
copies of the Software, and to permit persons to whom the Software is
10+
furnished to do so, subject to the following conditions:
11+
12+
The above copyright notice and this permission notice shall be included in all
13+
copies or substantial portions of the Software.
14+
15+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21+
SOFTWARE.

0 commit comments

Comments
 (0)