-
I think it might be interesting to find, if there is any, correlation between what _the each layer is trying to capture or how strong is each neuron activation when the model is given a non modified i…
-
Hi guys,
I've trained a Faster-RCNN model and attempted to visualize it. I have no problem in visualizing the first stage of the model; however when i tried to visualize the second stage layers, I …
-
I train a htc model,and I set test_cfg.rcnn.score_thr = 0.2.
Then I test the model, but in the result file results.pkl.json
I found some bboxs who‘s score below 0.2, is any thing wrong?
here is my…
-
I have trained the SSD three times with the default "ssd312_coco.py", but I got the ap with 21.4% or 17.2%, which was much worse than the result reported in the website. I noticed that the final train…
-
## ❓ Questions and Help
I download models from the modelzoo, but I want to use them in caffe2 for inference. Can I convert the .pth file to .pb file?
-
To train Cascade mask rcnn and the config file is following:
```
model = dict(
type='CascadeRCNN',
num_stages=3,
pretrained='open-mmlab://resnext101_64x4d',
backbone=dict(
…
-
i want to use the fasterRCNN_FPN to do detection. But the num_class of pretrained model on coco dataset is 81, mine dataset just has one class, And it throw an exception.
File "/home/ed/anaconda…
-
Hi!
I know this is not exactly your job or even what issues are about, but I'm super interested by your work, and I tried to understand the inner workings of the library. So I stumbled on the spati…
-
I really love the feature of the [3d Texture Synthesis](https://colab.research.google.com/github/tensorflow/lucid/blob/master/notebooks/differentiable-parameterizations/texture_synth_3d.ipynb), and I'…
-
#### Issue Description
It is currently not possible to convert a MultiLayerNetwork into a ComputationGraph in a straightforward fashion. My use case is the following: I've got a class which contain…