This project provides an official implementation of our recent work on real-time multi-object tracking in videos. The previous works conduct object detection and tracking with two separate models so they are very slow. In contrast, we propose a one-stage solution which does detection and tracking with a single network by elegantly solving the alignment problem. The resulting approach achieves groundbreaking results in terms of both accuracy and speed: (1) it ranks first among all the trackers on the MOT challenges; (2) it is significantly faster than the previous state-of-the-arts. In addition, it scales gracefully to handle a large number of objects.
MIT License
163
stars
21
forks
source link
Need help in figuring out the proper configuration for training #15
self.parser.add_argument('--data_cfg', type=str,
default='../src/lib/cfg/data.json',
help='load data from cfg')
self.parser.add_argument('--data_dir', type=str, default='/content/FairMOT/src/data')
On running I get the following error
json.decoder.JSONDecodeError: Expecting property name enclosed in double quotes: line 7 column 5 (char 101)
Because I only want to train and test on mot20 and not on other datasets I removed the part about other datasets from data.json file
Furthermore should the data_dir be set to where MOT20 is or should it be set to data in src directory.The same applies in data.json
As mentioned in the readme,I have downloaded the MOT20 dataset and organized it as follows
Then inside gen_labels_20.py i have mentioned it as follows
Then finally I changed the dataset directories as mentioned in
the data.json looks like this
And in opts.py the changes are as follows
On running I get the following error
Because I only want to train and test on mot20 and not on other datasets I removed the part about other datasets from data.json file Furthermore should the data_dir be set to where MOT20 is or should it be set to data in src directory.The same applies in data.json
Will appreciate it if you could clarify