Meet Patel
Meet Patel
For the loc_loss calculation, absolute difference between ground truth and anchor have been taken instead of normalized difference between ground truth and anchor. Why is it so? Am I missing...
I have created two files. One for label and one for prediction as per Readme file. For these predictions I have used iou_thresh=0.4 and conf_thresh=0.1. Now my query is if...
Have you used the pretrained MobileNet model and construct yolo layers on top? And used pretrained MobileNet weights for Mobilenet initialization? Or you just took the architecture of MobileNet and...
There should be provision to change the opacity of the boxes. This is useful when we have multiple objects in the image overlapping over each other.
It would be very helpful if we get to select the version of Tensorflow for which the results are displayed. There have been some major changes from Tensorflow 1.X to...
### Feature detail Google Photos and iPhone Photos have a feature where photos are organized based on faces found in the image. Also, there are some self hosted solution providers...
Hey, you have showed such an amazing work to train NNs in C++. I would like to know what are the reasons for which you started training models in C++...
I didnt understand why we need to tile the input in routing function? Please provide some clarity. Ref: [routing function](https://github.com/bourdakos1/capsule-networks/blob/84eb67a5b56456fc0a24d7fed8b0a53982fbd1c2/capsLayer.py#L121)
In ANN and CNN, the weights are not dependent on batch_size. But here ([wij](https://github.com/bourdakos1/capsule-networks/blob/84eb67a5b56456fc0a24d7fed8b0a53982fbd1c2/capsLayer.py#L122) and [bij](https://github.com/bourdakos1/capsule-networks/blob/84eb67a5b56456fc0a24d7fed8b0a53982fbd1c2/capsLayer.py#L92)) it is tiled according to batch_size and that number of weights will be trained....
As per your instructions, I have helen images and labels in .pts file for each image. I dont know how to prepare rectangle.csv file and what will be the data...