I find that when it comes to cross domain, osnet_ibn and osnet_ain work quite well compared with other models. But in general, person reid is just a more difficult problem to solve.
Hi, thanks for your explanation. This is really awesome. I have go through the GitHub repo. I didn't find any way to infer reid on real time i.e. on a video. There is only I find is training and evaluation on some already designed datasets. Can you please help me out with this?
Hi, if you use reid every single frame, it can hardly be real time. To achieve real-time performance, usually it is a combination of reid and tracking, as tracking is much faster than reid, while reid is more accurate.
All images from multi cameras are stored in the gallery, and when we query a person, we try to find the same person in the gallery with different camera id.
@@codingcat3274 That's very slow, because I was thinking of using it for realtime. Example: when a person leaves one FOV and moves to another FOV and performing re-id in such cases..do you think it would be application in such cases?
Yes, I think that is what person reid is designed for, but to achieve real time, there might be sacrifice on the model accuracy and you might use some approximate search algorithm.
@@codingcat3274 Thanks for your insights on this, would you mind recommending some good algorithms for aproximate search? I am working on this as a part of my Thesis I have to build a system
I can send you the ranking visualization results via email. But I highly recommend to run the torchreid repo since there are so many combinations of various datasets / networks / hyperparameters.
It only performs person reid. You could use some naive tracking frame to frame. But when there is occlusion, it is difficult for both tracking and person reid.
That depends on your training set, the number of epoches varies according to your training set size. For me, I usually start with something around 40 and then check the curves, if it already converges, try to make it 1/2. But I will keep it bigger than 10.
Thank you for this amazing presentation! I would like to integrate person re-identification algorithm in OSSDC VisionAI platform, see my latest videos with demos. Do you have an example on how to run this on webcam video source?
I checked your demos. They are really cool! You can have a gallery dataset, and preprocess the images to get embeddings for each person id with the person reid model. Then in your query video, each person is detected by person detector first, compute the embedding according to the person reid model, and compare this query embedding to gallery embeddings. And you can re-identify the person by finding the shortest distance in gallery images and a threshold. It is very similar to face recognition.
@@codingcat3274 So for huge shopping malls/public places wouldn't this approach be too slow as there will be thousands of gallery images and we need to process all those for each frame
@@SAINIVEDH Right, usually, person reid model won't be used for every single frame, usually it is a combination of person reid + tracking. For example, every 10 frames, use person reid model once. Or when there is great occlusion or the person quality score is low, use reid model.
![The moment we stopped understanding AI [AlexNet]](http://i.ytimg.com/vi/UZDiGooFs54/mqdefault.jpg)
![เจ็บที่ยังฮัก - ม่อน วรวิทย์ [ Official Mv ] จอนนี่มิวสิค](http://i.ytimg.com/vi/KALIGSRSjFs/mqdefault.jpg)
Thanks For the Video🙏🏻 very much appreciated!!
Metrics Explanation is spot on
Glad you enjoyed it!
great presentation, i would like to see more videos on prson re-id if possible, thank you in advance
Can you make some further videos on person reidentification?
This video is super helpful! could you share some of the code you used to run the experiments?
I mainly use this open source code: github.com/KaiyangZhou/deep-person-reid. It is pretty handy.
Why no one uses OSNET. Isn't that architecture specially designed for Person re-id tasks ??.
I find that when it comes to cross domain, osnet_ibn and osnet_ain work quite well compared with other models. But in general, person reid is just a more difficult problem to solve.
Great Video. Thank you.
Hi, thanks for your explanation. This is really awesome. I have go through the GitHub repo. I didn't find any way to infer reid on real time i.e. on a video. There is only I find is training and evaluation on some already designed datasets. Can you please help me out with this?
Hi, if you use reid every single frame, it can hardly be real time. To achieve real-time performance, usually it is a combination of reid and tracking, as tracking is much faster than reid, while reid is more accurate.
how you make multi camera to communicate so that each camera predict the same id.can you please explain it
All images from multi cameras are stored in the gallery, and when we query a person, we try to find the same person in the gallery with different camera id.
Great presentation, may I know how much time does it take for inference on image pairs?
I am considering this with Yolo+Tracker(Possibly opencv one).
Depends on the gallery size if you want to find all image pairs, but for getting the embedding itself, it takes about 200 ms for each image.
@@codingcat3274 That's very slow, because I was thinking of using it for realtime. Example: when a person leaves one FOV and moves to another FOV and performing re-id in such cases..do you think it would be application in such cases?
Yes, I think that is what person reid is designed for, but to achieve real time, there might be sacrifice on the model accuracy and you might use some approximate search algorithm.
@@codingcat3274 Thanks for your insights on this, would you mind recommending some good algorithms for aproximate search? I am working on this as a part of my Thesis I have to build a system
You can check out ANN for searching.
Helpful video for me.. Thanks 😀
Glad to hear it :)
what is the possibility to provide your snipped demo? Thanks in advance.
I can send you the ranking visualization results via email. But I highly recommend to run the torchreid repo since there are so many combinations of various datasets / networks / hyperparameters.
i want to ask that this torchreid model does perform tracking as well with the reidentification or it just performs the reidentification?
It only performs person reid. You could use some naive tracking frame to frame. But when there is occlusion, it is difficult for both tracking and person reid.
thank you for the resource shared
Glad it was helpful!
Thanks for the presentation. Is there any way that I can have your slides?
Sure, I can send you via email.
@@codingcat3274 here is my email thanh1705 at hotmail dot com
@@codingcat3274 Hey! Thanks a ton for the video. Could you share the slides with me too? My email is varunsdeens98@gmail.com
I am interested in '' Person Reidentification '' research, so if you can guide?
The topics were randomly chosen. But I could prepare more videos on Person Reid.
@@codingcat3274 I will be waiting for it! 😊
@@codingcat3274 hello, i'm also interested in person re-id, i'm waiting for your new
Can you run Bag of Tricks and A Strong ReID Baseline source code?
Yes, I could, the results are based on running their code.
@@codingcat3274 can I get your weight, 120 epoch too much
That depends on your training set, the number of epoches varies according to your training set size. For me, I usually start with something around 40 and then check the curves, if it already converges, try to make it 1/2. But I will keep it bigger than 10.
@@codingcat3274 how your mAP, is it bigger than 80%?
@@codingcat3274 my train 25 epoch 71.4 mAP, second train 58 eooch 70 mAP. Look not good
👍Thank you
Glad it helps :D
thanks
No problem!
Thank you for this amazing presentation!
I would like to integrate person re-identification algorithm in OSSDC VisionAI platform, see my latest videos with demos.
Do you have an example on how to run this on webcam video source?
I checked your demos. They are really cool! You can have a gallery dataset, and preprocess the images to get embeddings for each person id with the person reid model. Then in your query video, each person is detected by person detector first, compute the embedding according to the person reid model, and compare this query embedding to gallery embeddings. And you can re-identify the person by finding the shortest distance in gallery images and a threshold. It is very similar to face recognition.
@@codingcat3274 Thanks!
I'll try what you suggested.
@@codingcat3274 So for huge shopping malls/public places wouldn't this approach be too slow as there will be thousands of gallery images and we need to process all those for each frame
@@SAINIVEDH Right, usually, person reid model won't be used for every single frame, usually it is a combination of person reid + tracking. For example, every 10 frames, use person reid model once. Or when there is great occlusion or the person quality score is low, use reid model.