haotian/duan8
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

add detr-res101 support (#645)

Browse Source 
* add detr

* Update README.md

* add int8 quantization

fix some known bugs

* add detr-res101 support

update detr README.md

update rcnn README.md
This commit is contained in:
JumpPandaer 2021-07-26 00:37:42 +08:00 committed by GitHub
parent f81d5ba7f4
commit abcc753b19
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 32 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

31
detr/README.md
View File

@ -42,11 +42,41 @@ sudo ./detr -d detr.engine ../samples
3. check the images generated, as follows. _demo.jpg and so on.
## Backbone
#### R50
```
1.download pretrained model
https://dl.fbaipublicfiles.com/detr/detr-r50-e632da11.pth
2.export wts
set first parameter in Backbone in gen_wts.py(line 23) to resnet50
set path of pretrained model(line 87 in gen_wts.py)
3.set resnet_type in BuildResNet(line 546 in detr.cpp) to R50
```
#### R101
```
1.download pretrained model
https://dl.fbaipublicfiles.com/detr/detr-r101-2c7b67e5.pth
2.export wts
set first parameter in Backbone in gen_wts.py(line 23) to resnet101
set path of pretrained model(line 87 in gen_wts.py)
3.set resnet_type in BuildResNet(line 546 in detr.cpp) to R101
```
## NOTE
- tensorrt use fixed input size, if the size of your data is different from the engine, you need to adjust your data and the result.
- image preprocessing with c++ is a little different with python(opencv vs PIL)
## Quantization
1. quantizationType:fp32,fp16,int8. see BuildDETRModel(detr.cpp line 613) for detail.
2. the usage of int8 is same with [tensorrtx/yolov5](../yolov5/README.md).
## Latency
@ -55,4 +85,5 @@ average cost of doInference(in detr.cpp) from second time with batch=1 under the
| | fp32 | fp16 | int8 |
| ---- | ------- | ------- | ------ |
| R50 | 19.57ms | 9.424ms | 8.38ms |
| R101 | 30.82ms | 12.4ms | 9.59ms |

2
rcnn/README.md
View File

@ -128,7 +128,7 @@ sudo ./rcnn -d faster.engine ../samples
## Quantization
1. quantizationType:fp32,fp16,int8. see BuildRcnnModel(rcnn.cpp line 276) for detail.
1. quantizationType:fp32,fp16,int8. see BuildRcnnModel(rcnn.cpp line 345) for detail.
2. the usage of int8 is same with [tensorrtx/yolov5](../yolov5/README.md), but it has no improvement comparing to fp16.