Home

Awesome

DAC 2018 Design Contest

For general questions regarding this contest, please contact Xinyi Zhang xinyizhang@pitt.edu.

Usage

The get started, users have to run the following command on PYNQ-Z1 board:

cd /home/xilinx/jupyter_notebooks
sudo git clone https://github.com/yunqu/dac_2018.git

Remember the user name and password are both xilinx for super user.

After the above step is completed successfully, you will see a folder dac_2018 under your jupyter notebook dashboard.

Folder Structure

  1. overlay: This folder stores the overlay needed for the design. Usually it includes <teamname>.bit and <teamname>.tcl.

  2. images: All the test images are stored in this folder.

  3. python: This folder contains the python classes needed for the design, as well as the example notebook(s).

  4. result: The results includes the coordinates of the resized pictures and the frame rate of each picture.

12/06/2017 update

Redundant variables in DACSDC.py were moved.

Function "storeResultsToXML" in DACSDC.py is revised.

Please ignore object "boat9" and "group4" in the training dataset. They will be deleted from the latest training dataset.

01/05/2018 update

Revised python classes and polished notebooks.

01/18/2018 update

Updated the overlay: (1) changed to use HP port for DMA transfer, (2) enlarged the FIFO depth, (3) updated the DDR memory parameters, and (4) updated DMA configuration parameters.