mrhosseini

Adding dilated convolution and support for reorg3d layer

1

I have added dilated convolution as discussed in #37. It works both for TensorRT and CUDNN. The code also adds `reorg3d` layer ~~as an alias for `reorg` layer~~. There is...

Bug: incorrect output dimensions in yolo layer

4

In yolov3_tiny network dimensions of the first yolo layer is `13 x 13 x 255` and dimensions of the second one is `26 x 26 x 255`. However, if we...

Problem in reorg layer

7

Using [this network configuration](https://github.com/ceccocats/tkDNN/files/4776580/config.cfg.txt) the output of reorg layer using CUDNN and TensorRT is different from the exported outputs of darknet using CPU: ``` == OUTPUT 28 CHECK RESULTS ==...

When channel is 1, draw_box() fails because its been written for 3-channel images.