haotian/duan8
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
396 Commits 2 Branches 0 Tags 7.8 MiB
ad2c7a14d0
Commit Graph

396 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Mehmet Onurcan KAYA
ad2c7a14d0
kNumAnchor correction (#1429) 
It was not working for number of anchors != 3
 2024-01-16 14:49:27 +08:00
Wang Xinyu
2df7dd7f91
Update postprocess.cu 2023-11-14 18:36:17 +08:00
Wang Xinyu
f39f2c1433
Update yolov8_det_trt.py 2023-11-03 16:18:36 +08:00
Wang Xinyu
06d3a40a78
Update README.md 2023-11-02 09:50:45 +08:00
lindsayshuo
1d3862d99c
Fix cuda post processing (#1395) 
* Fix cuda post processing

Fix cuda post processing

* Update postprocess.cu

---------

Co-authored-by: Wang Xinyu <shaywxy@gmail.com>
 2023-11-01 15:00:13 +08:00
lindsayshuo
2fafde1af6
Fix the issue of int8 quantization compilation errors in YOLOv8 (#1393) 
* Fix the issue of int8 quantization compilation errors in YOLOv8

Fix the issue of int8 quantization compilation errors in YOLOv8

* Update model.cpp

* Update model.cpp

---------

Co-authored-by: Wang Xinyu <shaywxy@gmail.com>
 2023-10-30 15:26:53 +08:00
emptysoal
120121891d
modify yolov8 cpu postprocess iou (#1390) 2023-10-30 15:15:21 +08:00
Rex
c997e35710
yolov8n (#1387) 
* 修改yolov8nchannel数量以及 coco_file引用传递

* 优化代码风格

* 优化代码风格
 2023-10-23 10:56:10 +08:00
Wang Xinyu
6a267e0413
Update README.md 2023-10-17 21:04:00 +08:00
Rex
9a53b7f226
Yolov8 seg (#1381) 
* v8_seg

* fix code style

* fix code style

* [优化代码风格]

* readme中增加yolov8_seg分割的使用

* 更新readme

* 更新readme

* [fix python infer bug]

* Update README.md

* Update README.md

---------

Co-authored-by: Wang Xinyu <shaywxy@gmail.com>
 2023-10-17 20:59:21 +08:00
Charles_Huan
7b95f981f6
Change resize to letterbox to keep model pre-processing consistent (#1371) 
* Update model.cpp

Fix class mode pooling layer

* Change resize to letterbox to keep model pre-processing consistent

Change resize to letterbox to keep model pre-processing consistent

* Update yolov5_cls.cpp

reuse the preprocess function

* Update calibrator.cpp

* Update calibrator.h

* Update yolov5_cls.cpp

* Update yolov5_cls.cpp

* Update calibrator.h

* Update yolov5_cls.cpp

---------

Co-authored-by: Wang Xinyu <shaywxy@gmail.com>
 2023-10-11 15:05:01 +08:00
Wang Xinyu
3271725dd4
Update README.md 2023-09-27 18:40:38 +08:00
Charles_Huan
67b5b7a9ea
Update model.cpp (#1369) 
Fix class mode pooling layer
 2023-09-27 18:37:55 +08:00
何恺悦
eef6c5a711
fix builder->destroy() error in TRT8.6 (#1368) 2023-09-27 18:30:17 +08:00
lindsayshuo
12c19acce8
retinafaceAntiCov tensorrt8 (#1366) 
* retinafaceAntiCov tensorrt8

* Update decode.h
 2023-09-27 18:29:18 +08:00
lindsayshuo
24811e0520
Fix network issues with yolov8n (#1365) 2023-09-22 17:31:20 +08:00
lindsayshuo
3cbb7cd257
Fix abnormal network structure issues in Yolov8n and add inference function for multiple batches (#1338) 
* Fix abnormal network structure issues in Yolov8n and add inference function for multiple batches

* repair function

* repair function
 2023-07-21 14:30:11 +08:00
lindsayshuo
ae3bd5e6b2
Move GPU post-processing separately to cuda_postprocess.cu (#1331) 
* Move GPU post-processing separately to cuda_postprocess.cu

* Update postprocess.cpp

* Update postprocess.cu

---------

Co-authored-by: wang-xinyu <wangxinyu_es@163.com>
Co-authored-by: Wang Xinyu <shaywxy@gmail.com>
 2023-07-18 13:23:56 +08:00
wang-xinyu
e9a972e332 add symlink 2023-07-14 17:04:08 +08:00
lindsayshuo
70d0adc9ec
Yolov8 adds cuda post-processing function (#1326) 
* Yolov8 adds cuda post-processing function

* Yolov8 adds cuda post-processing function

* Yolov8 adds cuda post-processing function

* Yolov8 adds cuda post-processing function

* Yolov8 adds cuda post-processing function

* Yolov8 adds cuda post-processing function

* Yolov8 adds cuda post-processing function

* Yolov8 adds cuda post-processing function

* Yolov8 adds cuda post-processing function

* Update postprocess.cpp

* fix user id

* Yolov8 adds cuda post-processing function

* Yolov8 adds cuda post-processing function

* Yolov8 adds cuda post-processing function

* Yolov8 adds cuda post-processing function

* Yolov8 adds cuda post-processing function

* Delete images

* Update postprocess.h

* Update postprocess.cpp

---------

Co-authored-by: Wang Xinyu <shaywxy@gmail.com>
 2023-07-13 13:13:56 +08:00
lindsayshuo
bc4f854dc3
Add an engine export that can customize category models (#1322) 
* Add an engine export that can customize category models

* Add an engine export that can customize category models
 2023-07-05 10:37:09 +08:00
Prince Patel
30ea76e395
Update README.md (#1298) 
gd gw are not required in yolov7
 2023-07-03 10:24:03 +08:00
Wang Xinyu
4c9ef43ba1
Update README.md 2023-07-03 10:11:40 +08:00
Wang Xinyu
be8f6dc66c
Update README.md 2023-06-30 15:03:23 +08:00
lindsayshuo
fee7cbd0fe
yolov8 (#1321) 
* yolov8

* Update README.md

* Update README.md

* Update main.cpp

* Delete yolov8/output directory

* Delete yolov8/output directory
 2023-06-30 15:01:55 +08:00
HaiyangP
877bb24389
Optional nms methods supporting for rcnn (#1274) 
* Support to output the instance segmentation results on an original input image, rather than on a padded one.

* modifications according to freedenS's suggestions

* modification for padding

* merge preprocessImg and preprocessImg_ and delete all the debug codes.

* bug of reference passing

* make the vars consistent with the previous rcnn.cpp and update the README

* support different nms methods

* cxx11

* add contributors and delete interpretations

* update README

* fix some problems

* fix the bugs

---------

Co-authored-by: benkang wang <benkang.wang@uisee.com>
 2023-04-19 17:00:14 +08:00
tsieyy
30de4562bc
Adapt ufld model to tensorrt8 (#1288) 
* make ufld adapt tensorrt8

* make ufld adapt tensorrt8 and tensorrt7
 2023-04-19 12:45:43 +08:00
nengwp
f2521a5ffa
RCNN upgrade to support TensorRT 7&8 (#1256) 
* rcnn upgrade to support TensorRT 8.

* Update RpnDecodePlugin.h

Remove Chinese

* Update rcnn.cpp

Remove Chinese

* Update backbone.hpp

Remove Chinese

* Update backbone.hpp

* Update rcnn.cpp

* Update rcnn.cpp

* Update rcnn.cpp

* Update MaskRcnnInferencePlugin.h

* rcnn upgrade to support TensorRT 8.x

* rcnn upgrade to support TensorRT 8.x

* Update macros.h

* Update README.md

---------

Co-authored-by: nengwp <nengwp@github.ai>
Co-authored-by: Wang Xinyu <shaywxy@gmail.com>
 2023-03-07 19:26:53 +08:00
nengwp
ef22b1d03b
UNet support TensorRT 8.x (#1245) 
* TensorRT 8.x supported

* Unet TensorRT 8.x supported, add Contributors nengwp

---------

Co-authored-by: nengwp <nengwp@github.ai>
 2023-02-28 10:23:40 +08:00
lishanxiuxiu
f92dcf43dc
Fix dead link in yolov7 readme (#1222) 2023-02-02 12:50:58 +08:00
wang-xinyu
c66b8b7e2c update readme 2023-01-17 15:58:37 +08:00
Wang Xinyu
208032970c
Refactor yolov5 (#1211) 
* refactor pre and post process

* move model, refactor plugin

* refactor yolov5 cls

* refactor yolov5 seg

* fix space

* rename samples

* update readme
 2023-01-17 15:42:47 +08:00
Wang Xinyu
57ba75f9a4
refactor yolov5_det.cpp (#1209) 2023-01-16 19:46:59 +08:00
mantuoluozk
f334a1aa38
yolop TRT8 support (#1202) 
* yolop TRT8 support

* yolop TRT8 support
 2023-01-11 14:05:42 +08:00
Prince Patel
772d353399
Draw label text for yolov5_seg (#1197) 
* convert class id to class name + added confidence + different colors for class ids already present + background box to match with yolov5 repo + added coco.txt

* Required changes: removed coco.txt and *.jpg files, updated readme

* Resolving comments

* Update utils.h

* Update yolov5_seg.cpp

* Update README.md

Co-authored-by: Wang Xinyu <shaywxy@gmail.com>
 2023-01-11 10:54:16 +08:00
Rex
7b79de466c
fix a problem in warmUp (#1194) 
Int process of warmUp，the pred is None, resulting in the error of "masks = self.sigmoid((result_proto_coef @ result_proto_masks.astype(np.float32).reshape(c, -1))).reshape(-1, mh, mw)"
 2023-01-03 15:58:16 +08:00
Rex
3c0f62e82f
delete useless code (#1188) 2022-12-23 16:54:55 +08:00
Rex
7caa7e0dd3
Yolov5 seg python inference (#1185) 
* add yolov5_seg_trt.py

* complete yolov5-instence seg

* fix yolov5-seg

* fix_yolov5-seg

* fix_yolov5-seg

* fix code style

* fix coding style

Co-authored-by: Wang Xinyu <shaywxy@gmail.com>
 2022-12-22 20:33:40 +08:00
Rex
d6a0b784c4
fix_yolov5 (#1181) 
* fix_yolov5

* add fix

* add Rex
 2022-12-21 19:40:23 +08:00
wang-xinyu
aac418e1f0 yolov5/7: fix batch processing bug 2022-12-21 11:06:35 +08:00
wang-xinyu
ebc54830ec refactor yolov5 dir 2022-12-20 13:42:59 +08:00
Wang Xinyu
fbad2261dc
yolov5: upgrade to v7.0 and support instance segmentation (#1177) 
* add seg

* yolov5 seg

* update wts and readme

* add todo

* update readme
 2022-12-18 23:15:07 +08:00
wang-xinyu
165b0a428c yolov5: rm imagenet_classes.txt 2022-12-16 22:02:10 +08:00
wang-xinyu
712a50bc08 yolov5: use downloaded imagenet lables 2022-12-16 21:40:19 +08:00
uyolo
90b2f1e445 yolov5: load strides from .wts file (#1172) 
* update load strides from model

* load strides from model file
 2022-12-14 21:08:45 +08:00
Wang Xinyu
cdf45858c6 update introduction 2022-12-13 23:31:44 +08:00
Wang Xinyu
cb9efbdf21 unet: fix bug and coding style (#1171) 
* unet: fix bug

* update readme

* remove useless code
 2022-12-13 22:37:41 +08:00
East-Face
f98182e1b1 UNet: adapted to Pytorch-UNet v3.0 (#1169) 
* add deconvlution

* Delete unet/samples directory

* Update README.md

Co-authored-by: liuxiangd <liuxiangd@mailyunda.com>
 2022-12-13 17:22:07 +08:00
Zeyad Mansour
edecefe4c7 Update main.cpp (#1167) 2022-12-09 21:04:37 +08:00
wang-xinyu
ad7e3052f4 fix yolov7x bug 2022-11-29 20:44:34 +08:00