TSD
TSD copied to clipboard
1st place models in Google OpenImage Detection Challenge 2019
Hi, I am trying to run TSD inside a docker using the following python (/lh points mapped the host directory) ``` from mmdet.apis import init_detector, inference_detector import mmcv import sys...
I have trained TSD on the Docker. - baseline: tsd faster rcnn: - dataset: stanford cars. envs: - cuda=10.1 - pytorch=1.3.0 - mmcv=0.4.3 My training log is as below, I...
How can i get the spatial sensitive heatmap ? Just like the bear.
When I try to convert pytorch model to onnx by running tools/pytorch2onnx.py, I encountered this error. **Script** ```python python tools/pytorch2onnx.py configs/OpenImages_configs/r50-FPN-1x_classsampling_TSD/r50-FPN-1x_classsampling_TSD.py checkpoints/r50-FPN-1x_classsampling_TSD.pth --out r50-FPN-1x_classsampling_TSD.onnx ``` **Environment** sys.platform: linux Python: 3.7.5...
I already modified the key of pretrained models, and load it into the model.
I tired mapping the output index predicted by the model to label name using a list created using this file (https://storage.googleapis.com/openimages/challenge_2019/challenge-2019-classes-description-500.csv). But, the ordering isn't correct. How do i get...
While running this, i encountered the error from mmdet.apis import inference_detector, init_detector, show_result_pyplot config_file = '/md/tsd/faster_rcnn_x101_64x4d_fpn_TSD.py' checkpoint_file = '/md/tsd/faser_rcnn_X101_64x4d_TSD.pth' model = init_detector(config_file, checkpoint_file, device='cuda:0') **Environment** sys.platform: linux Python: 3.6.9 (default,...
Thanks for sharing your great work! I wanted to know - have you shared the pretrained weights for your prize-winning OpenImages network with 600 classes? Per my understanding, the current...
As I understood, the TSD fine tunes proposals from the RPNs existing in Faster RCNN, Mask RCNN, Cascada RCNN etc. In RetinaNet we don't have region proposals but instead the...
I am trying to migrate TSD to mmdet2.0, anything is ok when training faster_rcnn_TSD only on sigle GPU. When I run TSD with DDP, some error happened. Similar error [#2153](https://github.com/open-mmlab/mmdetection/issues/2153)...