To receive notifications about scheduled maintenance, please subscribe to the mailing-list gitlab-operations@sympa.ethz.ch. You can subscribe to the mailing-list at https://sympa.ethz.ch

Commit 44dfebca authored by Carlos Guija's avatar Carlos Guija
Browse files

Update README.md

parent 00a54d92
...@@ -46,10 +46,11 @@ Now that the dataset is created, you can proceed and train the network. Make sur ...@@ -46,10 +46,11 @@ Now that the dataset is created, you can proceed and train the network. Make sur
Execute the following: Execute the following:
``` ```
python3 device.py create --device=nameOfTheDevice python3 device.py train --device=nameOfTheDevice
--dataset=pathToDataset --dataset=pathToDataset
--weights=coco --weights=coco
--video=video_nameOfTheDevice.avi --video=video_nameOfTheDevice.avi
--file=tracked_pts_nameOfTheDevice_nameOfTheMethod
``` ```
This will start the training of a model based on the coco weights. A folder will be created with the device name and a timestamp in the logs folder. Every epoch a intermediated weight file is generated. This will start the training of a model based on the coco weights. A folder will be created with the device name and a timestamp in the logs folder. Every epoch a intermediated weight file is generated.
...@@ -66,4 +67,4 @@ You can now proceed to evaluate a new video and set the weights flag = True in t ...@@ -66,4 +67,4 @@ You can now proceed to evaluate a new video and set the weights flag = True in t
> N.B. Our models were trained using 15 epochs and 100 steps each. You can always adapt to your needs. in line 76 and line 223. > N.B. Our models were trained using 15 epochs and 100 steps each. You can always adapt to your needs. in line 76 and line 223.
> Also, a potential optimization is image augmentation, to maximize variation. This procedure was not tested in this work. However an possible implementation is shown in line 210-220. > Also, a potential optimization is image augmentation, to maximize variation. This procedure was not tested in this work. However an possible implementation is shown in line 210-220.
\ No newline at end of file
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment