Closed RomeroBarata closed 6 years ago
In addition to this, can you give pointers to detect keypoints and track them for a single video? Thanks.
Ok, I have written a script which can now detect and track for a single video, without evaluating it. I can share it if lot of people require it.
I think that it would be great if you could share it @yannadani!
Please feel free to open a pull request if you'd like! Thanks!
It will be amazing @yannadani!
Cool. I will submit a PR around this weekend when I'm free.
@yannadani Hi, I am really interested in how to detect and track for a single video, without evaluating it, can you share the scripts? Really appreciated!
Hi, I have questions about how to do real time tracking. According to my understanding, we should first detect all of the persons in the image set, storing the model into the pickle file called 'detection.pkl'. And then we run the tracking code and store the tracking result into 'detection.tracking.pkl'. It seems like the two things can not happen synchronously. Is my understanding wrong? Thanks a lot.
Yes they can. The tracking can be done online, with the detector running frame by frame, and the tracker linking the tracks to the previous frame.
Does anyone know where is code for "detect and track for a single video, without evaluating it".?@yannadani, could you please share your code for detection and tracking without evaluation.
Thanks.
Hi,
Is it possible to extract the detected key points and tracking information without evaluating performance of the model? I'd like to use Detect and Track on a data set for which I don't have the ground-truth annotations. Thank you!