-
I have been using the Mask_RCNN algorithm. It works fine. I have followed the process of training and validation followed by testing. But now, I would like to change the Backbone, ResNet101 to a new …
-
- 负责人：
zhaoyuchen
- 测试环境
- 显卡驱动:418.39
- CUDA 9.0 CUDNN7
- 当前性能对比
CUDA90
backbone：resnext
场景 | Paddle | pytorch | 对比
---|---|---|---
单GPU | 2.601 | 3.035 | 差于 14%
backbon…
-
This has a very good explanation!!
Does any of you donne this without mask, I mean a new branch but with the mask remove just bouding boxes?
-
Hi everyone! Thanks for the awesome work and resources. I have been trying to use Matterport's Mask_RCNN for a tiny project. I'm running this on GPU but model.detect is taking upwards of 30 seconds to…
ghost updated
5 years ago
-
What changes in source code needs to be done to implement multiple classes? Default source code is only for single class. Kindly help.
Thanks
-
Hi All,
have few questions regarding usage of mask Rcnn for small applications...I am asking these questions because it seems pretty slow and needs lot of memory to process.
1) Is it possible to u…
-
使用的配置文件是co_dino_5scale_vit_large_coco.py这个配置，test_cfg的配置我把score_thr设置为0.9仍然没什么用，推理出来的框有1000个，就是设置的max_per_img=1000，好像score_thr和nms没有发挥作用？
```
test_cfg=[
dict(
max_per_img=1000,…
-
In the [Faster R-CNN pre-trained model](https://www.arcgis.com/home/item.html?id=8c0078cc7e314e31b20001d94daace5e ) for extracting windows and doors, how to extract high-level feature maps from the in…
-
Hello! In your four detection experiments, I noticed a significant difference between the reported result and the test result for `mask_rcnn_swin_small`. The reported result is 44.2 mAP, but the teste…
-
i got torch version 1.13.0+cu116,
mmaction2 0.15.0
mmcv 2.1.0
mmdet 3.3.0
i always meet this error with anyother version:
user@u…